When I read the part about somebody complaining about "illegal benchmarking methods", I had to sign up, JUST to complain about the person who would make such a silly comment. Then, I saw snakeoil's name involved in the comments section and him throwing around a bunch of FUD and using the word illegal down there as well.
At that point 1+1 suddenly became clear and I realized that the plague of stupidity known as snakeoil is as much of a problem on this site as he is over on the techreport where I frequent more often.
Knowing that now, I want to apologize for any time the author of this update spent addressing that moonshot complaint as it's obvious you didn't know what you were dealing with. snakeoil is known on many forums for being a FUD spewing fanboy with delusions of grandure. Don't waste any more time worrying about comments about "illegal benchmarking procedures" when that little snot wouldn't know what the term "legality" even means.
Looking forward to further updates. And the numbers in this one are exactly as expected, no surprises for anybody with REALISTIC expectations.
needs to much voltage and temperatures are almost 100 degrees centigrade.
''Cons: 1) Lotsa of voltage increase needed to go past 3.6
2) no good coolers *yet* (I purchased coolermaster hyper 212 plus, better than stock but....)
Other Thoughts: Everytime i get a processor from newegg its like i'm destined to get the bottom of the barrel. after seeing initial reviews from various websites i was thinking 4ghz would be easy to obtain. Well i hit 3.6 ghz on stock voltage stable with prime 95, took 1.336 volts to get 3.8 ghz, and finally took me 1.375 volts to get to 4.0 ghz. Could run prime95 stable @ 4.0 ghz temps hitting mid 90's but i guess i'll back off 4.0 ghz til i find out more on this processor. initial reviews stated you wanted to stay under 1.4v, but i'm iffy on that if i want this thing to run for 6 months. kinda wish i woulda opted for the i920, i recommend anyone does if that can get a d0 stepping.''
Comparing a product just off the assembly line with a d0 stepping in an 11 month 'mature' process...? Of course the latter will typically have much more headroom due to its refinement (ie, what is your point?)
As for newegg, they're the same as any other vendor. You either get retail or OEM, and I highly doubt Intel or AMD are sorting lesser binned parts specifically for them. The fact that newegg often has parts in advance of other retail outlets might suggest that you're often choosing earlier samplings from the production line though?
And I agree with the below poster, your crossover from TechReport.com isn't going unnoticed.
A full review of overclocking the lynnfield processors on stock voltages would be extremely useful.
A couple of questions for the reviewer:
Is it possible to set the Cpu mulitplier of the Core i5 750 at higher than 20x, 22x for example?
How high can the memory multiplier be set with a Core I5 750 processor?
And lastly, in the previous review of the lynnfield processors you mentioned that the PCIe controller on Lynnfields can't be supplied by power seperately, and that you need to increase the processors
vcore on high clocks to achieve stable overclocking.
My queston is: If we can set the multiplier at 22x and and FSB at 160MHZ, "160 x 10 = 1600MHz frequency of the memory", aka rasing the multiplier and keeping the base clock as low as possible, will we be able to achieve stable stock voltage overclock?
I agree about the 'how an average end user would use it' perspective, and have posted such elsewhere (where we were discussing Lynnfield benchmarks and this site's were included.)
I find it kind of important that Lynnfield seems to require added voltage to do even modest oc, since added voltage affects temperature and life expectancy of the CPU. From what I understand this is due to the on-die pcie controller which isn't separated from the rest of the CPU "voltage wise" and needs more more voltage when overclocking. Anyway, do you at present have any indications if/when this is going to change in future revisions?
We have a test on a particular SW, still it's interesting.
In a clock to clock comparison at 3.8Ghz thanks to a good cooling, K10.5 is 16% slower than Nehalem without HT, 34% slower than Nehalem with HT.
This meas that, under this SW, the Nehalem cpus are now pretty limited to stay in the power limits of 95 W/130W from the cpu power management. The 3.4Ghz Phenom has not problems with it's own 140W power budget.
The upcoming Intel 32nm cpu line looks a killer !!!!
I noticed that the P55 chipset has an integrated Intel Ethernet MAC, yet all of the boards I have seen so far make use of a 3rd party (i.e. Broadcom or Realtek) chip.
So can somebody guesstimate what an i7 920 overclocked to 4Ghz(45C) on air compares with among the i5's overclocked or not?
The i7 920 is $200 at Microcenter. For the person which doesn't overclock anything ever, I can see how this article applies. But for people that build their own systems, I'm not seeing the big bonus here.
The comparable i5, which I haven't heard good things about it's 4 core overclock, is running at 20 bucks more. And the P55 motherboards ain't any cheaper than the x58 yet.
So, in summary. What is the advantage of buying an i5 anything rather than an i7 920 clocked at 4Ghz on air? That's it. Thanks.
For those consumers that DO NOT overclock (and they buy MOST of the chips), the i5 is faster than the 965 out of the box.
For those who DO overclock, two things are known:
- The i5 can OC to the same clock as the 965, and
- At the same clock, the i5 is faster than the 965.
It sure is not clear, especially because we were talking about Intel processors, but I think that by 965 he meant the Phenom 2 variety. Which is why we got a very interesting -NOT- rebuttal from snakeoil.
ok people, i have read most of the comments posted by both the obviously rabid intel fans and the equally rabid amd fans and there is one thing that absolutely must be acknowledged...intel's turbo feature IS overclocking, plain and simple.
overclocking means to run a processor at a higher than rated speed by either raising the fsb (even with cpu's that have integrated memory controllers the bios still lists a fsb for simplicities sake) and/or increasing the default multiplier.
with the core i5/i7 we have just that: a dynamic increasing of the clock speed from the rated default speed and said increase is only activated when the cpu is under load.
the core i5 750 has a rated clock speed of 2.66 ghz but the fact of the matter is that under most conditions it will be running at 3.2 ghz, it's absurd to compare it to a core 2 or phenom 2 running in the 2.66 ghz range and proclaim the i5 750 as the superior chip when the 750 is benefiting from 600 mhz clock speed advantage.
it reminds me of the muscle cars of the 60's when car companies would purposely under state the horsepower ratings by measuring horsepower and torque at a point below the rpm range when peak power was made, so that it could look like their cars needed less horsepower to be faster than the competition.
in fact, as late as circa 2000 chevy released a camaro ss that was rated at 305 horsepower that was faster than the corvettes of the day, hell the damn thing was ripping a 13 sec flat 1/4 mile time and anyone that ever drove one could tell you that it made way more than 305 hp. one magazine did put it on a dyno and found that it made about 350 hp at the wheels which works out to about 380 or so at the fly wheel.
same thing with intel: they market a chip as being a 2.66 ghz part and then under any condition than idle have it run at 3-3.2 ghz, my question is why not just clock it by default at 3.2 ghz, sell it as such and simple allow it to be down clocked when idle to save power, just the way they have done with the penryn and conroe and the way amd does with it's processors.
turbo mode is a marketing gimmick, nothing more and if i did buy a i5 750 (and i'm seriously considering picking one up) i would turn turbo off, buy a good after market cooler and overclock the chip myself to about the 3.5 ghz range, which should prove to be a very stable over clock.
"overclocking means to run a processor at a higher than rated speed"
"with the core i5/i7 we have just that: a dynamic increasing of the clock speed from the rated default speed and said increase is only activated when the cpu is under load."
Higher than the default speed, yes. Higher than the rated speed, no.
Overclocking means running the chip at a higher speed than the manufacturers settings. Variable clock speed is not overclocking. A chip at stock setting are never overclocked.
That said testing maximum overclock at similar cooling situations is also interesting. What is never interesting is comparing chips at worse than their out of the box settings.
"the core i5 750 has a rated clock speed of 2.66 ghz but the fact of the matter is that under most conditions it will be running at 3.2 ghz, it's absurd to compare it to a core 2 or phenom 2 running in the 2.66 ghz range and proclaim the i5 750 as the superior chip when the 750 is benefiting from 600 mhz clock speed advantage."
This is true, it should be compared to chips of the same price and power consumption. And this comparison is generally favorable to the i5.
"they market a chip as being a 2.66 ghz part and then under any condition than idle have it run at 3-3.2 ghz, my question is why not just clock it by default at 3.2 ghz,"
Because that would allow the chip to run at these speeds at full load on all cores, and they do not want the power/heat to ever raise to those levels. The beauty of turbo mode is that you maxize the performance withing the set power/heat limit. No competing chip do that, overclocked or not.
What's Montague? It is nor hand, nor foot,
Nor arm, nor face, nor any other part
Belonging to a man. O, be some other name!
What's in a name? That which we call a rose
By any other name would smell as sweet.
So Romeo would, were he not Romeo called,
Retain that dear perfection which he owes
Without that title. Romeo, doff thy name,
And for that name which is no part of thee
Take all myself.
This is for all you idiots that keep talking about the term overclocking, like it's a meaningful conversation.
I'll agree turbo mode is a stupid name, since you're not using forced induction, but the technology itself is valid. Idiots can argue over the term, but it gives one a finer granularity on increasing clock speed than a simple overclock would. It's intelligent in the sense it overclocks based on load, and it's a perfectly valid and useful feature for the vast majority of people buying the Lynnfields. They aren't going to overclock their processors in the conventional sense, and get an out of the box experience that gives them extra performance, without violating thermal and electrical characteristics of their processor and system. It's a good technology for a lot of people. Call it whatever you want. That's not the important aspect of it.
I agree with most people here that it's pretty useless, since most of us would just get an i7 920, or, for those who don't want to spend so much, get a i5 750 and overclock it. But, most people buying computers don't do this, and Intel makes most of their money on these people. It's a really good technology for them.
Totally agree its great for mass market level (turbo mode).
Virtually noone who buys a dell desktop is gonna want (or know how) to overclock.
I simply disagree that this makes 90% of the i7 line redundant, especially when youre talking to a tech audience. To an enthusiast (and most people whore going to buy x58 are going to be this level) the 920 is still imo, the best buy *in most cases*.
I completely agree. I'd much rather have the i7 920 than the brain-damaged Lynnfield platform. But, let's keep it in perspective. Say I don't have that much money to spend, and my choice isn't an i7 920, then the i5 750 comes into play. Only when the Lynnfield isn't competing with the Bloomfield does it make sense for the overclocking crowd. But, there are some price points the Bloomfield can't reach, but the Lynnfield can. At that point, you have to compare it against equal cost platforms, and in that context it could be an attractive processor even for people that are comfortable overclocking.
But, again, I am one of the biggest proponents of the i7 920 on these forums, so much so that people twist my words on it. Outside of my instinctive dislike of absolutes, I would say the 870 and 860 are essentially worthless processors for the technical savvy, as is the i7 950 (I still can't bring myself to putting the i7 before the brain-damaged Lynnfields). But, as easy as it is to say you're better off spending the extra $100 or $150 for the i7 920 rather than getting saddled with a castrated platform, that's an entirely specious argument if the person buying it simply can't pay it.
I would say the biggest problem with the i5 750 is the lack of IGP to go with it. It's a Celeron, without the platform to go with it. In the broader market, this might prove to be a bigger impediment to widespread adoption than the emasculation the chip suffered from. While you can add a discrete card, it adds cost, chews up a slot, and probably generates more heat. It's not horrible, but not ideal for the target audience.
"I would say the biggest problem with the i5 750 is the lack of IGP to go with it. It's a Celeron, without the platform to go with it."
If you believe that a $200 processor is a Celeron then you seem to have little knowledge of current market conditions. That's without considering the excellent performance of Ci5. Clarkdale is what you are looking for, not that anyone will call Clarkdale a Celeron, except perhaps those that look down their noses on dual-core.
"the core i5 750 has a rated clock speed of 2.66 ghz"
The i5's rated speed is 2.66 stock. 2.93 4 cores active, 2.93 with 3 active, 3.06 with 2 active and 3.2 with 1 active. That's what it runs at "stock" with default BIOS settings. I don't know (or care) whether it's OC'ing or not, I just know that it runs at those speeds by default. And it does so in a very effecient manner.
The turbo isn't a gimmick. If it was I wouldn't be using it.
I will be buying i5 and Noctua cooler. If I was buying i7 or Phenom II (which was a big possibility) I would be overclocking it to 3.6-3.8 24/7 stability permitting.
With i5 turbo on, I can overclock to 3.4, consuming less power and still reach 4Ghz in single core apps. While running COOLER, and less VOLTAGE then the Phenom and i7 alternatives at this same speed. It handles all of this on the fly. Without needing different OC profiles for different games or apps. Do you know that when only only 1 or 2 cores are active the heat and voltage necessary is alot less? That's why the single core apps can run at 4 Ghz safely. Unlike the Phenom or i7, which can only get to 4Ghz while OC'ing all cores instead of 1. It's the same reason why you can OC a Phenom II 550BE higher then a 955BE. And why when you unlock the extra two cores on the 550BE, you have to reduce your OC.
This isn't a gimmick this is effeciency at it's best. I'm sure in the future there will be better solutions, but for now I am delighted with this. It's a powerful single core, duel core and quad core all in one. It's like a 550BE + a 955BE. You get the strengths of all of them with none the weaknesses. Since I am primarily interested in gaming, the lack of HT isn't an issue. And the AMDs don't even have HT anyway. With this, "turning turbo off to OC" becomes a thing of the past. Unless you like ineffeciency.
This i5 will be my first Intel CPU since Celeron 766mhz.
"It's a powerful single core, duel core and quad core all in one."
That's a great way to describe Turbo mode. It ends the question of whether to get a faster dual core or a slower quad core, since you can get both. I suspect you can also force the CPU to stay as a faster dual core by disabling 2 cores at the OS level.
I'm going to explain you the real story behind this intel turbo mode.
intel was having a problem with core i7 920, this processor was the only affordable core i7 but it was killing the more expensive core i7 (950,975) because can be overclocked to 4.0 ghz and obtain the same performance for less money even when it has a locked multiplier.
we need to kill overclocking before overclocking kills our profits intel thought.
So intel decided to create the turbo crap story, or auto overclock.
Instead of giving overclocking as a free bonus we will create this turbo feature and begin charging for it.
that's true intel is now charging their users for the overclocking which with amd phenom 2 is free.
So intel is using all the available overclocking the processor may have and call it turbo (like turbo time in toy story).
This way they can segment their lineup, how? because they make a model overclock 600 mhz, and they can make the turbo more or less aggressive depending on their marketing needs.
Turbo mode is just overclocking, is not fairy dust, is hocus pocus
intel uses to skin some intel ignorant users.
Intel never liked the idea of giving overclocking for free, now has decided to kill overclocking because turbo auto auto overclocking consumes all available overclocking headroom, and intel can use it to charge more for a processor and to segment properly their lineup.
In the end it doesn't matter if Turbo is overclocking or not. This is a silly argument about semantics. What matters is that in their stock configuration it's fair to compare the Lynnfield with other CPUs also in their stock configuration. Which means that it's reasonable to compare an i5 750 (stock, i.e. Turbo is left on) with an equivalent AMD processor. Question is: what is an equivalent AMD processor? DIfficult to say, really.
Where I live, for example, the 95W version of the 3GHz 945 is priced similarly to the 750 (it's actually a bit more expensive, but I'm guessing that this might be different in other countries), so that would be a good candidate. Similar price, similar thermal enveloppe, similar frequencies (remember that Turbo will push the frequency of the 750 between 2.8 and 3.2GHz).
But I am tempted to say: who cares, apart from the fanboys?
It does matter if reviewers are going to completely dismiss CPUs with no or less aggressive turbo modes.
The inital review said that there was no point in any i7 below the 955. This is clearly an absurd assertion.
It is basically just because of the higher turbo mode enabled results taht anand has made this statement. As weve discussed turbo mode is really manufacturer endorsed overclocking (with slightly more sophistication) so surely the review is flawed as youre not comparing like for like.
I take it you're one of the second group of people who take exceptions with the article: no not the AMD fanboys but the proud owners of a Bloomfield, probably a 920 - or about to get one? What's wrong? Overnight it didn't become a bad CPU. It's still a fantastic piece of silicon, I'd say. Anyone who says the contrary is deluded (snakeoil anyone?). What one need to understand is that Anand's statement implies something along the lines of: with all things being equal (like retail channels maturity, which at the moment gives Bloomfield an edge) and for prospective buyers that don't need the 32 lanes of PCIE, etc...
Why get so upset?
So-called 'manufacturer endorsed overclocking' has very little relevance. It's semantics. AMD is doing exactly that with their Phenom2s. Ridiculous voltages that pushes the thermal enveloppe. So what? There are 2 states for a processor: its drop-in stock state and its altered (either manually or via an auto mobo thing), i.e. OC'd state. What's wrong with comparing CPUs in their stock config? WHat's wrong with comparing OC'd CPUs at similar frequencies? I don't see it.
Oh so were going for the "you must be an XXXX" tactic..... Implied bias if you cant actually argue the point. Ok then.
No, actually I run multiple PCs, a Q6600 a 920, a C2D and a phenom 2 for a variety of purposes (server, gaming x2, work and multimedia), I also do some system building (and pre sale overclocking). I dont intend to change machines personally in the next few months so consider me an interested observer.
What I object to is simply the supposedly unbiased reviewer making such a retarded statement.
I totally agree that the new chips are pretty good. However the review makes it sound like theyre a huge leap forward when infact theyre simply intel realising they have headroom in their architecture.
Hell, Im much more interested in the integrated PCI-E than the gimmick of turbo mode.
As for the semantics thing. Yes it is semantics. Hence why Im bothered about anand making a such a song and dance about it. The new i7s are damn good cpus but to suggest the performance increase is due to anything other than the "overclock" of the aggressive turbo mode (which is something attainable EASILY by the old i7s) is totally fanciful.
Look, I totally agree an aggressive turbo mode is a good thing for the average consumer. The merits of turbo mode are not in debate. Its unequivocally a good thing for the average buyer.
The ONLY thing I disagree with about the review is the massive amount of guff thats been put out making out that turbo mode is a revolution. Its really not.
Its quite simply intel realising the headroom they have in their architecture and cleverly doing something useful with it (coincidentally exactly what overclocking also does).
The assessment of this is what I disagree with.
When a review site is aiming at a tech audience (like anand does) surely there has to be some differentiation between the audience and the average consumer? If the review had said something like "For the non technical user the new CPUs offer much more than the current i7 lineup" Then Idve totally agreed with it. As it was the review uses the emotive and totally wrong assertation that "there no point in any i7 below the 965". Thats just stupid.
Imo actually the real situation at enthusiast level is that there is NO justification (unless money is genuinely no object) in any i7 except the new CPUs (for people who dont know how/are unwilling to overclock) and the 920 (as it pretty much always clocks to similar levels as the 965 anyway).
So, to me, anand totally missed the point (given the audience) with the inital assessment. This is what disappointed me, this is what Im bothering to argue about.
And for the record its highly ironic that the people calling others "fanboys and morons" tend to be the ones who dont back up theri points. Take a step back, THINK and perhaps youll see that some of us are actually talking sense rather than blindly reading a review and cheer leading.
Not quite sure why that was a reply to me. Apparently you agree with what I said. You called Turbo Mode a gimmick. Now you say it is not a gimmick. End of story.
So, before we could selectively increase the clock on particularly cores according to core loading conditions? Perhaps looking at the Xeons will help you understand. One model is a 45W TDP model with a 1.86GHz clock. However, it has the ability to Turbo past 3GHz, all within the specifications of the processor.
I don't know if your last paragraph was aimed at me, but I didn't call YOU a fanboy. OK maybe categorizing you as a 920 owner might have been a 'cheap shot' but it sure did hit the target.
Now to come back to matters at hand. Nobody denies that the 920, 750 and 860/870 are based on the same micro-architecture. And Gary's preliminary results of these CPU OC'd at the same frequencies show near identical performance. But the main difference is that now we have a mainstream processor as opposed to an enthusiast, which means that in a lot of cases it's going to be left in its stock state: large OEMs like Dell or HP, people who can't or won't OC, etc... And for these situations, it seems that for example, the 860 is a better processor than the 920. Sure it's a close match but at least the 920 keeps its dignity. Imagine if Intel had come out with the 860 at 3.06GHz with no Turbo in 3/4C and just a 3 Turbo in 1/2C. It would have been a massive slap in the 920 owners face. Also the Turbo allows Intel to keep the TDP to reasonable levels. The 860 is probably more often than not a 3.06GHz (minimum) processor but you can't call it that because it will throttle back to 2.8GHz in some circumstances. And it's OC'd performance is almost identical to that of the 920 (it would appear). Again there are some cases where you are better served by the 920. Clearly. But the fact remains that soon enough you will have combos mobo+860 a good deal cheaper than 920+mobo and for 'better' performance. Hence Anand's statement.
I know this might come too late for this specific review, but maybe you guys could do this one later.
Okay, most of the reviews seen for the i5/i7 cpus are compared to amd's top of the line Phenom II X4 965 BE, while this is perfectly valid to compare top of the line cpu's, it's also a 140W processor, and the power consumption tests show that, also the board used to test said cpus is also a top of the line board which in my opinion doesn't compare well with P55 based boards and their moto.
So what I'm asking is if it's possible to compare offers from AMD and Intel in their usual configurations as in, mid-range.
This would probably be a 785G board paired with a 65w (P2 x4 905e/x3 705e) or 95w processor (P2 x4 945/P2 x4 720).
For the intel camp, this would be a P45 board with the usual suspects (up to the Q9550 and their low power versions if you have them around).
As for the new brothers in arms, the i5 750 and i7 860 would be my choices (even if the 860 has a current asking price a bit out of the league of the previous cpus, it's the only new one with ht enabled).
Don't take me wrong, I, like most of the readers, like to see new hardware pushed to the limit, but at the end of the day my personal computing goals are much more mid-range oriented and I just don't see justification in passing the monetary range of the choices I listed above.
Think of it as AT for the masses of poor economy people that still want something new :P
Saw the drama queens throwing their hissy fit over how things "should" be tested. Those clowns need to realize they do NOT speak for most users, they're just a vocal minority. I skipped most of the remaining comments so sorry if I missed something.
I'm building two new systems, one w/ a i7 920 and another with a either an i5 750 or i7 860. My lady and I want these specifically for gaming and plan on doing mild OCing w/ minimal voltage increase (preferably none, but looks like p55-based cpus demand at least a tad extra voltage). Gaming benchmarks showing this scenario are helpful and I look forward to updates/articles comparing p55 versus x58.
More wish list: with the lowered power draw of p55 setups, I'd love to see what's being sucked outta our wall outlets under OC'd/SLI/Xfire setups since this can mean an extra ~$50 for a beefier power supply.
It might have something to do with the PCIE. As soon as you raise the BCLK, it overclocks the PCIE too which requires some voltage to stabilize. More than the actual processing cores it seems.
Clock for clock the i5 is faster. That is why I chose it before i5 was even released. The only advantge Phenom II has is 16x CF (but no SLI). No price advantage either.
If you OC both to 3.6 or 3.8 Ghz, i5 wins. Unless it's a game/program that favours the AMDs in which case, the Phenom will beat the i7 as well.
Stock clocks are most favourable for Phenom since it's 3.2Ghz vs 2.66Ghz. But that's irrelevant for me, because whether I get Phenom/i5/i7 I would OC to 3.6/3.8 with Noctua cooler. When they are both are 3.6Ghz, not only does the i5 consume less power but it's also faster, especially in CPU orientated tasks.
Since turbo is so effecient, I plan to OC to 3.5Ghz with turbo on, and that will give 4Ghz in single core apps, 3.9 in dual core apps etc. While saving power at the same time. Fantastic!
It's a no brainer choice for me, and I'm not an Intel fanboy (I use Athlon currently), I just want best bang for buck.
I'm looking foward to the followup when it is posted to confirm my thoughts.
Actually you can get a 780a or 790a Nvidia motherboard for 16x-16x SLI. Since the PII 955 BE can also hit 3.8GHz when overclocked, there's no reason to pay the extra $50 for the 965.
Agree about the 955. At the shop I will be buying from online in my country (cheapest in Australia) i5 is $285, 955 is $271. $14 difference. Gigabyte p55 UDP5 board is same price as the 790FX AMD board. For me the extra performance and effeciency (heat/power) is worth the extra few pennies.
Unfortunately regarding the AMD SLI mobos's, this shop doesn't carry them. But even if they did, it's the same problem. SLI only. It's either CF only or SLI only. If you are going to go the AMD way, your probably better off with the full Dragon platform with the AMD software to take advantage of it.
What I haven't seen any reviews talk about is this:
What if I want to have 2 x Video cards, along with another expansion card that uses PCIe 1x or 4x? Is it just not possible with P55?
Agreed, although dual graphoics cards don't interest me personally knowing how moterboads assign PCIe lanes depending upon slot population is important. If you have two 1x devices and one happens to be in the secondary graphics slot, does the primary graphics slot get cut down to 8x? It probably wouldn't be a huge performance hit but it's nice to know this type of detailed information.
So when's that mobo round up gonna happen? I've been refreshing the page all day! I thought you said Thursday/Friday we were going to see the first article, and then Monday the last? No rush... just can't wait to read it! :D
What about the mobo roundup from X58? We got like 3-4 mobos in a "roundup" when Toms used 3-4 SERIES of roundups each with 8 mobos (over 2 dozen mobos total).
What about that i7 overclocking guide that never came? What about SSD stuff? Huh? The DDR3 roundup?
If you go back article by article a lot of them promise MORE content "later this week" or whatever. Nothing EVER comes. I wouldn't get my hopes up about AT articles anymore. They're quality when delivered but if it's a "preview" with "more to come this week" or a "lab update," I wouldn't expect much at all.
Is the clock for clock comparison, with an average overclock. This represents a huge deal for me in comparing processors' actual raw speed compared to just wildly varying clock levels.
As part of your decision process I think you should note that this particular test, whlie no doubt one of just many, was of a pretty well threaded application. Turboboost *still* provides a notable benefit in even well threaded apps. Turboboost is more complex than you implied, just check the main review it's more than single-core oc'ing on demand.
What's fair and what's not leaves a lot to interpretation it seems. To me a fair comparison would be:
1- 965 stock vs 750 stock (ie Turbo is on because that's the way the CPU is regardless of the motherboard). People who won't/can't change BIOS settings. Let's say they bought a DELL, for example.
2- 965 3.8GHz vs 750 3.8GHz (Turbo off to get there). Gary did it. Check.
3- 965 Auto OC vs 750 Auto OC (Turbo off). WHy not.
4- 965 Best OC vs 750 Best OC (Both on air, closed box). Similar to 2-, wouldn't you say?
The point is, there are a lot of ways to do these comparisons, and Gary is doing a pretty decent job to cover most of the bases IMHO.
What happens if you are running a nVidia mobo?
See, Turbo mode is not like a regular oc. It does not affect other component, power consumption, heat, etc. It really does not have the drawbacks of oc. I thank Anand for explaining it. If you want to do an OC some of the P55 boards come with utilities that will oc the system by a small percentage at windows start up. This will be equivalent to the AOD.
You are seriously retarded. Please go play in traffic. Turbo is not overclocking, and your ignorance is getting really tiring on this matter.
"Overclocking is the process of running a computer component at a higher clock rate (more clock cycles per second) than it was designed for or was specified by the manufacturer"
Did you completely miss the part where he explained why single cores clock higher than quad cores? It has to do with the amount of voltage that can be used. If you have a process that isn't using all four cores, why use them? So shut them off and pump the voltage through to the other cores. This is not overclocking. This is meant to happen by the designers, creators, and manufacturers.
As a corollary to your last paragraph I think we can conclude that snakeoil's brain, in addition to being a low neuron wet state device in the first place, permanently has turboboost turned off as well so most of his brain is permanently shut off and he can never perform mental processes well.
"It automatically allows processor cores to run faster than the base operating frequency if it's operating below power, current, and temperature specification limits."
Intels marketing department says its not overclocking (yay) so people who actually believe everything is true in a marketing speil get suckered into it.
Turbo mode is just manufacturer stamped overclocking with some bells and whistles. Snakeoil is actually right but guess its easier to flame him because of the other things hes written.
So much missinformation around on overclocking (increase thermal envelope, increase power consumption etc etc) most of the points only apply IF you overvolt. Which you really dont need to do on most i7s.
But hey overclockings bad mkay (unless its written in a fancy marketing name on the side of the box, in which case its revolutionary).
That would be a neat trick: increasing clockspeed on all 4 cores, thereby doing more work per unit time, yet consuming no more power because you didn't overvolt? So much misinformation around on overclocking.....
I suppose it's what you mean by 'negligible' then. That carries a diferent connotation for processors than it does nuclear power stations.
Yes, turbo mode increases clock speed on the cores in question. What is the point you want to make? It is all done within the specifications of the processor. It's just a clever application of the idea that if all the cores are not working hard then there is room to clock some of the cores higher. If you want to call that clever overclocking then you can, but what most people mean by overclocking is exceeding the specifications of the processor.
A dynamic OC which does not change the thermal envelop. 965BE has 140W TDP at stock. When you overclock the the thermal envelop skyrockets, putting extraordinary presure on the power circatery. Turbo does not do that. Yes, it increases the clock of the individual cores, but does not put additional pressure on the system, it is the default setting, it is not user-initiated. It is the stock clock. It just varies. The clock is not set in stone. If I go and change internal clock then it is OC. It is over the clock it was meant to be.
Jon Stokes at Ars Technica noticed something odd in the last graph at http://techreport.com/articles.x/17545/6">http://techreport.com/articles.x/17545/6 . Basically Far Cry 2 at 1600x1200 4xAA is GPU limited (expected), but the Phenon II x4s and Core 2 Quads all converge on a frame rate that's approx 10 fps higher than the i5-750/i7-870 (unexpected).
The E8600 even beats the i7-975 at 1600x1200 4xAA, even though Far Cry 2 is obviously cpu limited at lower resolutions.
Any chance you guys could try to reproduce/investigate that, and look at single gpu resolution scaling to at leats 1920x1200.
I know that's the sort of thing you mainly look at with new chipsets, but with the northbridge and cpu merging i5/i7-8xx is effectively a new chipset.
I have spent the better part of the last twenty four hours trying to figure out our FarCry 2 numbers and it also happens in other games by the way. ;) The only conclusion I have right now is that with the latest NV drivers, Win7, and certain games, the Phenom II 965 is faster now.
Not only against Lynnfield but also against Bloomfield in the tests we utilize for the motherboards. The CPU oriented action test we use in the processor benchmarks tells another story. So something is happening along the graphics path at this point, what exactly is something we are trying to isolate.
Ah thanks. Good to know that you're looking into it. Does it seem to happen in many games, and is it limited to NVidia cards + win 7 only?
Since you can reproduce it, it's probably a good idea to make a note/show some benchmarks demonstrating it in an article soon, before too many people start replacing their existing Phenom II/Core 2 Quads looking for better framerates at high resolutions.
After reading this update and the comments, I went back to peruse the comments on the launch article. Ugly!
Granted, there is something to be said for improving comparisons at equal clocks, etc., but I can't help but get the feeling we are ignoring the writing on the wall. In short, the desktop PC (not servers or workstations) is rapidly going the way of the dodo. The last remaining performance-demanding app on the PC is gaming, and PC gaming is dying in its present form.
So the home PC trend is toward a console/nettop in the living room and maybe a NAS for local storage, synch services for phones and note/netbooks, tied to cloud backup. The P55 platform is a clear step in the direction of SOC, enabling Intel to stay relevant eventually in the living room (notice the high % of uATX boards, too). When and if Larrabee comes into play as a competitive graphics player, we might see an eventual renaissance of PC gaming, but on what has virtually become an Intel console, not a desktop. This is why P55 makes sense to me; it forks the market so that X58 and so on forms the backbone of servers and workstations, and P55 drives in the direction of consoles.
Unfortunately for NVidia, I do not see a space for them in the midterm, barring some miracle alignment of the stars.
Microsoft Flight Simulator is a great tool for testing these multi-core setups. The latest service pack includes the ability to use multiple threads, and has support for both DX9 and DX10.
There are many flight sim people out here that want to know how these systems will fare with FSX and x-plane.
In your review the multi-gpu sli and especially crossfire didn't work well at all with Lynnfield. However over at Tom's hardware it didn't seem to work all that bad. Do you have any idea why?
Is it the on-die pcie controller which causes problems or the shared x16 line? I would expect the x16 to be a bottleneck at very high resolutions but are there other scenarios it will cause problems?
Keep in mind, 2 card sli and Xfire scale fine on 2 x8 lanes. Three and 4 card setups are where the PCIe bandwidth is not going to be up to snuff. If I recall... the tests were shown with 2 4870x2 cards, with a setup like that you would need full x16 lanes, therefore a LGA 1366 setup would be needed.
So hypothetically speaking, I could clock it to 3.4 and when turbo kicked in it would throttle a couple of steps up to 4.0, maybe. That would be sweet.
Preliminary results from Anand shows a best-case OC (with Turbo on) for the 750 of 3.2GHz and with Turbo it would reach 3.84GHz on a 2C/1C stress (although Anand shows 4.16GHz for that last number. I'm still waiting to hear from him about it). Go check the main P55 article, in the overclocking section.
i see it's listed on MSI's website now, and people have removed the cooler to discover a hydra chip. any word on if this will be benchmarked in the mobo lineup, or is it reserved for its own multigup lineup instead? and are there any other motherboards known to have this chip? really need to know ASAP, and i need multigpu scaling with NF200 as well, cause unless it's good enough (comparable to x58 performance with 5870x2+5870 or equiv usage of bandwidth), i need to get my hands on an D0 920 before intel stops making them so i can be ready when i get my 5870 to replace my UD3P lol
i should also mention, this is with the 860 costing $30 more than the 920 at microcenter right now, and the fact that the S1156 boards with the PCIe lanes i need being at least $180-200 out the door, and ram being negligible since i would be selling another system when i get this one anyway. the 920 would be easier to overclock using less voltage, i dont care about the PCIe latency reduction since i would be using a bridge chip anyway, and im a RAM whore and an overclocker (4.2ghz quad and rising) so the S1366 is more ideal anyway in most ways. im just really eager to see if the hydra is all that it's cracked up to be, and i wanna see some OC results on the P55 FTW vs the EVGA X58 micro and standard (as far as ease of OC and max OC goes).
At the moment it looks like the 920 is a bit cheaper than the 860 so I think it's normal that people should question the wisdom of i7 1156 as opposed to i7 1366. But the 920 has been around a while so the retail channel has matured (lower prices, availability, stepping). When the 860 has reached the same level of maturity (if it does), I'd be surprised if the 860 is not cheaper than the 920. Anyway that could be a moot point because by then socket 1366 may have gone 6-core.
Thanks for the additional tests. I didn't bother posting in the comments of the launch article as they were already extremely long, but I am also one of those looking for clock-for-clock comparisons. As the application I would be designing for is Photoshop, I'd want HT. This puts the i5 out of consideration for me, and as the platform cost of an i7 860 isn't a lot different than a i7 920, I approach it more along the lines of "Prove why I shouldn't just go straight to the 920 and keep the door open for future higher-performance options". For either one I would apply some easy overclock ( I was thinking more like 3.3GHz) and would be happier if the motherboard could do this automatically. I'm basically coming from more of a computer-as-a-tool than computer-as-a-hobby perspective, so I look forward to the new results.
Since my previous message may have gotten lost in the flood of flamewars, I'll just briefly state it here again, seems appropriate.
I'd like you guys to also spend some time on reviewing the onboard audio features, particularly since there seems to be a new generation of VIA onboard chips. It's hard to get any info on what these are really capable of, and how mature the software/drivers are that ship with them. It could mean the difference between having to get an addon soundcard or not, when buying a motherboard.
I will have a short synopsis on the audio choices provided on the P55 boards in the motherboard articles. Based on the updated VIA chipsets and Windows 7, we are thinking right now that a separate piece on the state of on-board audio would be wise. I will see what we can do to get that done this month. In the meantime, I like the VIA 2020 better than the Realtek ALC 889a from an overall perspective, the drivers still need a little work but audio quality is better in my opinion.
That would be nice, Gary.
One feature in particular that has made a big difference for me, was the support for realtime Dolby Digital Live encoding. When Realtek started offering this feature on their onboard chips, it meant that you could get digital 5.1 output to any Dolby-capable amp/home stereo. You were no longer depending on the analog outputs, which required lots of wires and generally was of poor quality with onboard solutions.
Later I got a Soundmax, which could do DTS Live encoding. The sound quality was excellent.
(I'm not talking about having stereo sound upconverted to 5.1, but about having 5.1 audio from games and such encoded in Dolby or DTS in realtime, and sent over the digital output).
So I'm particularly interested in what the VIA chips can do in that respect.
Just remember that there's a difference between judging chipsets and implementatins of said chipset. Especially for analog use there will be variation among implementations.
Now comes the fine grain; This is where subjective opinion comes into play. What is "reliable" ? Apparently Subjective. What is too much attention on a per name brand basis ? Apparently Subjective.
Let us toss out a fictional example.
Lets say product name A get a lot of attention, and has a strong "fan base" Product A is typically known as being the fastest on a per product test basis, but in the IT sector is not known as reliable. Product A gets tons of raving reviews despite this, even despite the fact that from time to time they also have some very serious problems. Product A does very well in the OEM sector ( probably with tons of OEM support from them working together as partners or not ).
On the flip side of things products name B has been in the business for an equal amount of time, or possibly longer. Product B is known mostly as a very reliable product. Product B is also nearly as fast as Product A ( some times even very rarely surpassing that of Product A ), and the people in the IT sector have no qualms putting this product into a personal system, or even in a production server. Product B is not without its flaws, but fewer by comparrison, and some serious flaws rarely. Product B has also pioneered several features that are still in use today by many product names, but is now out of the retail sector after one such serious flaw, and a bit of financial troubles.
So, if "review site A" gives raving reviews of a product name continually in favor despite the fact that this product will not run countless hours/days/months without crashing ( because this is not important to them ). Then of course review site A is going to get some flack for their very public words. However, ANY OS can be very, very stable with the right hardware. Yes this even includes Windows. It will probably never be perfect, but solid hardware, with solid software backing it can make a huge difference. I've been there. am there, and hoping to continue to be there in the future.
Sadly for review site A however, I take their words with a grain of salt, and have to back whatever is said on this site with countless hours of research on my end. That is ok though, I do not pay review site A for anything, and they are still very informative even if in not exactly the same way I wish.
You've certainly piqued my interest: Where can I read about Intel instability in server environments? Being a mere mortal, I only know what I read at hardware review sites like Anandtech, so I haven't heard about this before.
CPU's very rarely exhibit serious issues in general, and if we're going to get into that Intel has always been very solid. This also is in the context of new parts.I am sure the mention of Intel as a specific brand in this blog was a generalization. As a matter of a fact, I am a long standing "fan" of AMD. I am glad AMD is around, and in in the past have used *many* AMD CPU's ( also Cyrix if you want to get into that as well ). Now days, I actually prefer Intel CPUs. The *only* problem I have ever personally had with Intel is that in the past, their CPUs were not obtainable simply because of the outrageous cost for many people. Now days, and probably within the last 4 or more years, their prices have been very reasonable. This is not to say that Intel as a whole is not without their problems. Read: the company, not the product.
The context of my post was more along the lines of motherboards, memory, power supplies, video cards etc. And while I left out details on purpose, the example was loosely based on real events. I bet Jarred knows which products I was talking about. Because I have talked about them more than once over the last several years. Also this example was an extreme generlization.
Actually, I'm not entirely sure what you're talking about, in part because you were so vague. My guesses:
Intel IGPs still frequently suck, especially in terms of drivers. NVIDIA chipsets also have issues on a regular basis... but then, problems with drivers and chipsets have plagued pretty much every company.
As a whole, I would choose Intel chipsets over the competition, but it's a tough call. (Note: if we're talking about IGP chipsets, that recommendation would change. Sadly, good IGP chipsets frequently come with the less desirable other items, like unstable chipsets/drivers or slower CPUs.)
While I have used plenty of AMD setups in the past, I buy for what is currently the best platform. That means all of my Athlon 64/X2 systems are gone now, except for the one Athlon 3200+ 939 setup I keep around for my daughter. My main work and play systems are actually both running Kentsfield processors -- Q6600 @ 3.30GHz and QX6700 @ 3.20GHz. There's also an old Pentium D sitting in the corner gathering dust... though it has an X38 motherboard so I should probably get a different CPU from Anand or Gary.
The rest of my computers are now all laptops in various stages of testing. Two are AMD-based, and frankly they're the slowest and least desirable *laptops* by a large margin. The Dell Studio 14z is actually looking like a very good laptop that starts at $650, and $200 extra will get you pretty much everything you really need. Battery life, performance, stability, and even a 1440x900 LCD, with a package that weighs in at less than 4.5 pounds. You can get AMD-based setups for less money, but they're slower in every area and offer about 2/3 the battery life.
Totally off topic post, I know, but some of you might be interested in reading it. :-)
Speaking of laptops, I purchased one recently too, and could not be happier with it. It is not a high performance jobby ( but I did look at one of those too; By Asus even ), but it also only set me back $399 , or $450 if you include the 4GB Crucial kit I purchased at the same time.
The main reason I purchased this laptop is that everyone I have spoken with that has owned one of this brand has been very happy with them. But these people also try to take care of their equipment, so they may/may not be bullet proof. The one I purchased is definitely flimsy, but I do not move it around a lot, and when I do, I go out of my way to take care of it. So, it is not perfect either, but I am very happy because it does the main thing I expect from any system I own, or build. It does not crash.
Speed is important too in some cases, but when you purchase a pre built system ( laptop in this case ) with a T3400, 2 GB of RAM, 160GB HDD, and Intel 4500M IGP. What are *you* expecting ? I would also expect AMD laptops to not be top performers, but you have to know this going in. This is why lots of "research" is nearly mandatory, other wise not matter what you purchase is going to be a big let down.
So, I guess my point here is that "desirable" in this context is subjective. Myself, I purchased a laptop mainly because we're 100% solar ( off grid ), and could not afford to spend $30,000 for batteries, plus more for other equipment like panels, or wind turbines *JUST* to run my desktop. By comparison, it is a beast, and is gaining on a couple of years old now. It also sucks down 224W continuous while playing Modern games. 165W while idling. The laptop ? 40W while gaming, or 17W idle. Big difference. So I chose an Intel laptop only based on they seem to be leading the market in low power systems that are still usable for most every day tasks. Read: an atom ain't going to cut it. That, and I payed less than what they seem to be asking for a high end atom system.
Strikeback hit it on the head. Asus, and ABIT. Roughly, very generalized, and blown out of proportion.
I will miss ABIT sorely, but not because of the name, because of the product. I had never, ever had any product from this company that was not maintained for a long time, or at least until all the issues were ironed out. I am far from alone in this school of thought. VERY far. The products in the end were perhaps a tad slower than some of the competition ( they usually landed in the middle of the pack from memory, with a rare speed demon from time to time ), but were very solid as far as Stability went.
They were not however perfect, as no one is. I had a system board go south after 5 years, because of the substandard capacitors used. 5 years . . . and they were not the only company who fell into this pit as it were. The board was a KT7A RAID-100 board, still have it ( for parts ), but it does not function any longer despite a cap replacement.
Anyways, my point on this subject is not to glorify ABIT ( although maybe they deserve it ), or blow smoke up anyones butt. My point is that the practices used by this company should be the rule, and not the exception. Instead, we get hardware that people from reviews call stable after running successfully for a very short amount of time( and in the real world the darned things will not run a week without some form of a crash related to hardware ). This is even with stock proper settings. And it is not even the reviewers fault, except that perhaps these products get awards that they should not have, until they exemplify true ability by reputation. Enter our "budget line" product . . .
Now, this is not just a problem with motherboards. This is a problem that seems quite capable of popping up in any part of the market. Many of us know this, and avoid these companies/products like the plague. This is not because of fanboiism, this is because those of us who do run small businesses do not wish to lose business because of substandard garbage. Maybe my thinking here is antiquated . . . but you know, if it "just works" I can actually get some work/play done, instead of fixing the damed thing ( again ).
By the way, I prefer systems with Intel chipsets, and if the system is for play, I just drop in a discrete card. I am not adverse to other products, but lately the product would have to be proven to me in person. I am also pro Intel CPUs, even though I have used many AMD CPUs in the past( even before the Athlon ). Cost factor is what played the biggest role for me, because like I said in a previous post. CPUs are very rarely an issue. Sometimes different branded CPUs are slower, sure. That is a given. If the CPU exhibits serious problems, hell we'll find out about it before we need to use one ( because of sites like Anandtech ), and we can just bite on the next revision. Right ?
Wow I didn't realize Abit was gone... just read the Wikipedia article. I was wondering why I hadn't seen a board from them in a while. Kinda sad. I used to like their boards back in the day. :( Though I did also have problems with one and that was what got me looking at other brands for my motherboards since.
The reasoning behind this is random writes on laptop drives are very slow. Random writes on some SSDs are slow too.
With the default NtfsDisableLastAccessUpdate setting, Windows will write to the disk every time it opens a file or directory.
I suspect this might affect performance a lot. Maybe you can benchmark it :). Naturally this might screw up some apps, but I have NtfsDisableLastAccessUpdate=1 for years and it's fine for my usage.
So basically, clock for clock these things run almost identical with LGA1366 processors. That's really interesting, but doesn't take away that as a stock processor the i5 750 makes a really sweet choice at a great price for those who don't want to overclock and want lower power consumption
Thanks for taking the time to do the extra tests. Good to know youre listening.
Yeah some people went over the top with the stupid irrational criticism but what did you expect with incendiary remarks like (paraphrasing) "all CPUs below the 965 are now pointless".
To me personally, that was what really annoyed me about the 1st article. A silly blanket statement ignoring many of the real world reasons for choosing a 920 or whatever.
Anyway, thanks for the additional testing. Lets hope thats the end of it regarding this debate. Anandtech has always been a site Ive admired for its high quality and insightful testing i guess to have read one review that I disagreed with in 3 years isnt really too bad!
It really boils down to the target markets. In the consumer space, the 1366 platform is a luxury and 1156 has proven worth the wait. That being said it was marketing genius to put out the top gun platform first and bilk the early adopters.
I commend the earlier poster for identifying his present needs, and contrasting that with the relatively short tick-tock production cycle Intel pursues. Any kind of long term investment in a desktop Intel chipset is a financial mistake now, and a technological mistake in 18 months.
Other market space could be interpreted differently...
Because it's the cheapest entry point for LGA1366/X58, which allows for a wider upgrade path than LGA1156/P55. E.g., memory (24GB vs. 16GB); ECC (possible vs. never); cores (6-8 vs. 4); PCIe lanes (36 in various flavors vs. 24 in few flavors).
Is that potential upgrade path worth the premium? Depends. If you're looking at longer term system-level price/performance, then arguably yes. I was waiting to see what the new parts offered before making a decision on building a new home server, and for me LGA1366/X58--at the cheapest I can get into it, which means a Core i7 920--is still attractive.
Curiously, I thought that too. However, the more I thought about it, the more I realized that they're not really points at all. I've had this Core2Duo since 2007, and it's performed exceptionally well for what I've wanted it to do. Now that I feel that I may want to "upgrade", my options are Core2Quad, 1156 or 1366. Well, The Core2Quad is more or less "useless" right now, given how well even the 1156 runs. So having the chipset (in my case, a P35) upgraded means nothing to me.
So now I look at upgrade time (it's been 2 years, so I figure it's about that time again). So my choices are the 1156 or the 1366. Do I really think that I'm going to need more than 8 gigs of RAM right now? No, I doubt it. And that's on the cheap side. I figure that when I'll need more than 8 gigs of RAM (12, 16 or 24 gigs), a 4 Gig stick of DDR3 will be "cheap enough" to be palatable. So the real question is, will I need more than 16 gigs of RAM? I honestly can't imagine needing to more than quadruple my current memory usage. I don't even use the 4 gigs that I have now. I think that I'd be happy with 2 gigs, now that I think about it.
So, "future proof" I think is a bit of a misnomer. How long can we reasonably expect the 1366 to be relevant in the Market place? My old P4 was one of the socket 478 chips. The "upgrade" for that ended with the P4, 3.06 GHz (with hyperthreading). Needless to say, it wasn't worthwhile for me to actually spend money on the CPU upgrade. I'd question the logic of making a "future proof" decision, at least based on CPU capabilities.
The possibility of 6 cores or 8 cores might be interesting. It might be interesting if we had a significant share of applications that took advantage of more than 4 cores (or 8 "virtual" ones). I don't have any that are appreciably good at that. Most of those (video encoding) will still be stunningly faster when run over a GPGPU calculation (you can't beat 800 stream processors, or whatever NVidia has right now).
That leaves just the PCIe 2 * 8x or 2 * 16x issue. I don't run dual graphics cards, so that option isn't very useful to me either. The idea sounded good that I could, in the future, buy a second card when my single card "got slow". However, I'd just as soon buy a nice, new card that performed on par with the 2x cards, so that buys me nothing. Particularly given that finding the same card would be harder to find since either the company has re-spun a new rev of the cards that aren't 100% compatible in SLI or Crossfire, or that they're just not made any more. So the cost of the card wouldn't be that much cheaper.
I dunno. I always thought that "future proofing" a purchase sounded good. However, I've honestly never actually taken advantage of that. When things get too slow (and more RAM thrown at it doesn't solve the problem), I usually just buy the new big 2 (CPU which usually begets a new chipset, hence also buying a new MoBo). Every now and then, I can use the RAM from one base computer to the next.
Appreciate the effort,any chance of maybe some old scores from a C2Q maybe even C2D etc for some comparisons, to see if it would be worth people updating their existing systems?
I second that motion. I too appreciate the efforts.
I would like to see some PCMark and 3DMark scores for an E6600 and a Q6600 included in these reviews so I can see how they compare to i5 and i7 LGA 1156 processors. Many of us mainstream people will be upgraded from these processors and want to know if its worth it. If that does not happen, then the only other alternative is to visit Toms Hardware CPU charts and scale with the i7 920.
Sometime between midnight and four am I plan on running a C2Q system for comparison. :) That delay is my fault. We had only planned on showing the P55 and X58 board results in the initial motherboard articles. We added the 790FX/965BE earlier this week and now if everything works out properly, we will have a P45/C2Q setup in the motherboard articles based on a large number of requests. However, for a quick CPU comparison, check out the Lynnfield launch article.
"I will interject a personal note here, the emails/private messages that outlined a strong case for additional research and testing certainly held a lot more weight than comments like "You are Intel's payroll...", "Worst review ever...", and the moonshot , "Illegal benchmarking methods..". First off, if we were on Intel's payroll we would not be working here (a logical conclusion right? ;) ) As for the other comments, everyone is entitled to their opinions. We do our best to keep an open forum and let the comments fall where they may, but offering constructive criticism and facts to back up those comments is what actually causes change, not endless shock posts or attention grabbing statements. I still have hope in people abiding by the rules of Internet Etiquette, but apparently we are still a long ways off from that happening. I will step off the soap box, well, until the next article...."
Gary, being a popular website on the net, it goes with the territory. Atleast with Anandtech, you are open and entertain other people's suggestions and opinions. I think that sets you apart from the rest. Keep up the good work!
The comment I made about this latest processor is, it's not a big leap by any means, it's really nothing to write home about and I felt AT was overhyping it a bit. I think the bigger leap (as I said in the other article) will be mobile Nehalem and beyond, and that will be an article worth reading.
I'm not sure if this will work here...but as a staff member at geeks 2 go we can set in our profile blocked posters ...people that we don't even see their posts...EVER...it would help keep the replies on topic or at a minimum not waste any time having to wade thru the BS posts by anyone we consider not worth reading anything they have to say...just my 2 cents worth...
You guys should just ban snakeoil. All he does is make outlandish and idiotic claims just to get a rise out of other commenters. He does this too at Techreport. Now he's going to think that he's justified in all the noise he tries to stir up. It really ruins the more intellectual atmosphere I've come to expect out of Anandtech comments.
Anyway. Personally, the test combo that I'm more interested in is testing with Turbo off as well as on. Since the performance boost provided by Turbo is variable based on whatever other apps the user or operating system may be multitasking in the background, testing with Turbo off provides the "guaranteed minimum", i.e. a baseline level of performance that a Turbo capable CPU will provide, and that any Turbo boosted extra is considered a free bonus.
Otherwise, people who skim through the graphs may think that the CPU will ALWAYS provide the Turbo boosted score, instead of understanding that the Turbo boosted score is an "ideal" score, when conditions of the CPU are at its best for Turbo.
"You guys should just ban snakeoil. All he does is make outlandish and idiotic claims just to get a rise out of other commenters"
Then they'd have to ban TA152H and others for the same reason. Just let them speak. I think it is an error for Anand or Gary to reply to them though. That just gives them status in their own eyes and encourages them to continue.
Actually, I back up my clains, unlike what you did. I don't just say this or that sucks, or was done wrong. I tell them why, and back it up.
Gary can complain about shock posts, but, I put more effort into my posts than they put into their P55 motherboard pictorals :-P. And I DO back them up.
Most things in life are not black or white, good or bad, but a shade of gray. People are simpletons, and like it simple, but it's just not how most things are. P55 isn't a perfect product, by any means, and it's not all bad. If you don't have people pointing out the flaws in the methodology, then you have holes in your information. Feedback is not only acceptable, it's necessary.
More to the point, when you idiots were all pointing out how good the performance was on the first P55 article, I was explaining the memory performance was abysmal for what the product was. No one else seemed to see it. Guess what? I was right. The numbers were improved for the released product (and I kept saying I thought it was from pre-release hardware), so the terrible performance didn't make sense after all. Did the writer see it? Nope. He was saying how it was the greatest thing since cheddar cheese? Did the masses see it? Nope. Like proper cattle that can't think for themselves, they just ate it up and did as they were told, so to speak.
You need critical people that are going to look at what you do, and take you to account when you do something wrong, or just don't do something. I need it when I do things to. I don't always like being shown it, but, at the end of the day, the end result is better.
That's really what they should be here for, right?
Actually I've hardly made any posts on Anandtech in the past. I've just read your particular brand of posting recently, since I was interested in the 1156 platform. You have said many incorrect things about the 1156 platform, both before and after release, using ridiculous, hyperbolic language to do so and what's worse, you never acknowledge your mistakes and habitually attack others. You are no better than someone like snakeoil. Actually, snakeoil may be wrong, but he is a bit less free in insulting others. I think you are funny sometimes though, just because you can be so wrong and technically incompetent. Writing this post, I still have to chuckle eg. about what you said about pin-multiplexing on the DDR3 and PCIe interfaces. I don't want to start a flamewar, so I won't reply any further. No point anyway.
I read stuff from little losers like you, and I'm wondering if you even have the common sense to realize the hypocrisy in what you're saying. Are you really that stupid? I guess you are.
You insult me, and then whine that I insult people? Duh. What a moron.
You say I was wrong about things, but I don't see a single quote.
You have heard about multiplexing pins right? Processors have done it before, in fact the 8086 did it with the address and data bus. Intel shaved off 210 pins. I haven't seen the pinout yet, but it's probably out there. Oh, and I never said they multiplexed the pins, I said it's a possibility they did, and it would be nice if we got information like that instead of pictures of motherboards.
There is no multiplexing of pins across the PCIe and DDR3 interfaces, I can tell you. I'm not going to waste time quoting you. Notice the only thing that could really be interpreted as an insult is that you are sometimes funny by virtue of being so wrong headed. Your mistakes, hyperbolic language, inabality to admit your mistakes, attacks on others or whatever are just facts that anyone can see.
So, I did reply further. But, what do you care, I'm a loser and a moron, right?
Just a further expand the point: you realise that the DDR3 and PCIe interfaces are separate interfaces, routed to different places on the motherboard? I was really laughing that you would even make such a speculation.
"Anyway. Personally, the test combo that I'm more interested in is testing with Turbo off as well as on. Since the performance boost provided by Turbo is variable based on whatever other apps the user or operating system may be multitasking in the background, testing with Turbo off provides the "guaranteed minimum", i.e. a baseline level of performance that a Turbo capable CPU will provide, and that any Turbo boosted extra is considered a free bonus."
Turbo off doesn't really give a "guaranteed minimum". If there is heavy mulitasking then performance on a given task can be much worse than the turbo-off score for that task measured in comparitive isolation.
Yes, it's true that extremely heavy multitasking can overload a quad core CPU with too many threads, thus degrading performance for all tasks, but that's not the fault of the CPU. That's simply more tasks than the CPU can handle. Also there's the side issue of disk contention as multiple threads fight for access from the hard drive. That could also affect a benchmark, but that's also not the fault of the CPU.
But I'm thinking of cases where the background multitasking pushes 1 or 2, maybe even 3 cores up to a constant and prolonged 100%, but the quad core still has enough "room" to theoretically run a single threaded benchmark without being affected by the background tasks.
For example, suppose an instance of Firefox and an instance of Openoffice get into a buggy state where each are running an infinite loop, thus is not causing disk contention. In this case, 2 of the 4 cores will reach a prolonged 100%, but there are still 2 cores available for a single threaded benchmark, such as Cinebench. On a Core 2 or Phenom II, there should NOT be a drop in Cinebench performance. But on a Lynnfield, Turbo either can't kick in, or will kick in with much less of a boost, so there SHOULD be a drop in Cinebench performance. Assuming all other conditions aren't otherwise affecting Cinebench, the performance drop should theoretically be no worse than a Turbo-off test. That's what I'm referring to about a "guaranteed minimum".
This is also why I suggested a multi-instance tests of single threaded benchmarks. That way, we can see how much of a drop off there is for 1x, 2x, 3x, 4x concurrent instances of single-threaded-Cinebench both in non-Turbo CPUs of Core 2 and Phenom 2, and also in the Core i5/i7/etc. Hopefully, Gary has read my suggestion and added such a test.
So, what you are essentially saying is that the cores have to be loaded just right: not too much and not too little, for the non-turbo score to be really representative of actual performance in a given task. For me, that's trying too hard to make it interesting. Not that I'm criticising your interest. I just don't really share it.
That's sort of what I'm saying. I'm interested in getting a scoring "range" that Turbo capable CPUs can provide to give a more complete picture. It's kind of analogous to how hard drives have a read speed range, example: 60MB to 80MB, depending on whether the data sits on the inside or outside areas of the platter. In normal usage, you're going to get various read speeds, but you won't read any worse than 60 MB/s, assuming no other external factors, but you also won't ALWAYS get 80 MB/s either.
Likewise, Turbo capable CPUs have a performance range from no Turbo to best possible Turbo. That's why I'm interested in the no Turbo scores, to get the bottom end of this range.
Look, you can get a range of performance from crawling along because of other things go on to performing at maximum with turbo-on (and reaching the maximum level of turbo) relatively isolated. I use my computer in the real world and don't choose combinations of tasks to make the turbo-off-relatively-isolated-figure especially significant. It might happen by chance that I get something like that sometimes, but it is no more interesting or significant than anything else.
Just about the only useful thing that sites can benchmark is the maximum performance that can be expected. It is understood that if you multi-task you can have a whole range of lesser performances for a particular task, depending on how heavy you multitask. The turbo-off-relatively-isolated figure is just an arbitrary figure plucked out of that range that might happen by chance sometime. It's pretty much a waste of time benchmarking it if the goal is to give people an expectation of what they will get in the real world.
That's a reasonable request, I'd be interested in seeing that too. Like Eeqmcsq says, Turbo is like a free bonus.
Also really looking forward to the expanded i7/860 tests. In the original article, it looked like a great balance between the pricey models and the cheap alternatives. It's a shame you didn't have more time with it then.
When looking at their review of the ASRock X58 Extreme I saw that in the comments snakeoil (the site's biggest AMD fanboi) and SiliconDoc (the site's biggest Nvidia fanboi) went at each other.
I am running the Turbo off numbers on the 750 today, which means getting these motherboard articles out is taking just that much longer. ;) However, we understand the importance of the request and I will show a few results late today. Anand will followup next week with a full analysis. Thank you for the suggestions.
Other suggestions I thought of that you might try for Turbo capable CPUs: Run a varied number of concurrent instances of a single threaded test. For example, run 1 instance of a single threaded test, then run 2 separate instances of the test at the same time, then 3, then 4. See what the numbers look like and how much Turbo contributes based on the # of busy cores.
You know, that's a really good point about a "guaranteed minimum". I hadn't thought of that when i defended the use of testing with Turbo on "because that's how it comes out of the box". I can appreciate your reasoning and agree with you that it would be helpful to also have Turbo Off numbers for the benchmarks. Looks like Gary will be doing that, so that's cool.
Gary, can you also add some results of OC with Turbo On. I know that the final OC will be lower, but in the initial review Anand pointed out that in a lesser threaded circumstances the effective maximum overclock with Turbo On is higher (i5/750 and i7/860) than the maximum OC with Turbo Off.
The benfits of running lower OC with Turbo On will not only be in single and dual threaded app's but also power consumtion => lower temps. Good for 24/7 OC.
I am intrested in the socket 1156 and I run my systems OC 24/7.
I have been working on that also. :) I just received another retail 860 today that on a clock for clock basis needs about 0.06V less than our first sample, so the results should be more in alignment with the retail processors on sale now.
Regarding Anand's OC numbers (with Turbo) in the main article, something is off for the 860 and the 750. I posted in the comments to the main article (around page 32/33!?!) about it. For example the Turbo 1C number would be 26 BCLK, ie a turbo of 6. Am I missing something?
If the rest of the results are like this one then I have a bad feeling the comments are going to get worse. These results do not bid well for those who asked for this information. Pulls up a chair and grabs the popcorn, just waiting on SnakeOil and the other nut jobs to show up.
I'm not sure 19 seconds is an annihilation, but at "stock" speeds the Phenom II 965 does do better than Lynnfield 750 at stock.
1) Aren't "stock speeds" determined, in part, by where the manufacturer sees there product's niche? In the past, parts of processors have been disabled for the very same reason. So it stands to reason this may be the case here. I guess we'll see. I recall a feature or two offered in the i7 series are not available in the i5. Again, I'm guessing it's marketing IMHO.
2) These tests determine a processor's capability. If you're complaining that we don't see "apples to apples" comparos between AMD and Intel, it's because Intel processors are more stable when over-clocked at higher speeds. So if AMD is competing with Intel for fastest processor, AMD is likely to set its "stock speed" closer to its stability threshold than Intel will with theirs.
So insisting that "stock speeds" are a relevant measure is like comparing two runners when they're jogging and determining who would win a given race. It's only when they're really pushing the limits of their abilities do we see who the best runner really is.
It is the same here. So based on the information we see, i5 clearly is a better performing processor for this application.
These arguments are only relevant to us system tweakers anyway. The average Joe couldn't care less. All he's interested in is realized speed and not the barely perceptible differences between two closely matched processors.
Could you explain to me how you call it fair when you compare
the Phenom II 965 BE which runs at 3.4Ghz (stock speed) to the Lynfield I5 720 that runs at 2.66Ghz (stock speed, turbo off) which
makes it 27% (740Mhz) slower in terme of clock speed?
If you take this clock speed into account (just to play fair like you want) then we should compare the PII 965 BE to a Lynfield I5 720 overclocked to run at 3.4Ghz also right?.
If you apply this 27% difference in clock speed to the Lynfield I5 750 (with turbo mode off) then I'm not quite sure that it's still slower than the PII 965 BE...(245 - (245 * 27%) = 178 compared to 188 on PII 965 BE).
Do you realize what you just sail OilySnake, you just said:
"our biggest most expensive chip can beat up you SMALLEST SLOWEST CHEAPEST CHIP"
Thats also like saying :
"My 12grader can beat up you're toddler"
Do you LIKE beating toddlers? is that what you are saying OILYsnake?
Your just as much of a troll and waste of skin here as you are on Techreport. Why don't you step into traffic and do the world a favor?
The tests are 100 VALID.
At stock settings, the tests reflect exactly what every intel user will get. if you dont like that.... stop reading the reviews you pathetic little meat sack.
or better yet, go design a CPU... that is if you can leave your mom's basement long enough....
OH and don't forget you IDIOT, you are the moron that was swearing up and down that I5 would not support virtulization...
tool.... no wait... your not good enough to be a tool.... your a tool-bag.
Anand, Gary.... ignore this idiot. talk to damage on Techreport... this guy is a complete douche.
This is stupid, The 750 with turbo boost is NOT an overclock. Its simply how the CPU manages thermal distribution. You may not be happy with the results. But until you can provide you're own data that is repeatable and doesn't cripple Core I5/7 CPU's, by say disabling turbo boost. Then you're just grasping at straws.
Face it AMD cannot currently compete in the high end. Get over it.
You make some pretty far fetch accusations, with no backing. At this point you're just trolling. It's to bad I fill the need to feed your ridiculous comments.
turbo is overclocking. if you are going to benchmark using an overclocked processor then that is cheating. and that's illegal
do you think people is stupid?
Let me just say that I've build quite a few of AMD's based system going back to the K6-2 era and they are pretty good processors, but your comment is just simply idiotic.
And I bet if AMD has this feature, you'll say it's a fair comparison.
true... i agree...
we don't hate AMD. they are pretty good CPU manufacturer.... but having said that, we are also end user... and all of us want the worth of our every single penny which we spend on CPUs. now there are ways.... poor people who can't afford a $280 cpu + $150 motherboard + $120 RAM and so on, will go for a AMD. it gives nice performance for its price. People who can afford that much cost will definitely go for intel.
At the end of the day, it is your pocket which becomes deciding factor rather than what really you want. I want to buy a Dual Xeon 5580 based system but can i afford it ?? NO. so what do i do ?? may be go for a Xeon 3550 or something... can i afford that too ?? no. so what do i do? may be go for a AMD 955BE or something....doesn't matter what i want. but if i can't afford it i'll choose the next one downside the list until i reach the affordable limit.
in current world situation, most people can afford a good $1000 system, since in that range, i7/860 gives best performance (no matter how that performance is achieved... be it OVERCLOCKING (according to you) or an innovation called turbo boost (according to rest of us)). the bottom line is what Gflops people are getting for the money that they are going to spend and as long as intel is fitting in that meter better people will go for it. when the time will come (and it has come in past once), people will go for AMD as well.
...people are stupid, maybe?
Btw cheating is not illigal. If it was, you could take AT to court and win (if they were cheating).
After all, do I need to remind you that the 965BE is an OC 955BE?
And do you really want to see OC (24/7 or max) 965BE destroyed by and OC (24/7 or max) i5 750.
I believe this statement speaks for itself. However, turbo is an innovation designed by Intel to increase performance. If that is cheating, how have we made it as far in the computer industry as we have? It is like comparing a single core to a dual core, and saying for the test to be fair that the dual core has to shut off one of its cores.
BTW, are you referring to the "turbo on" as auto OC? In that case you want to compare a Stock 965 to cripled (worse than stock) i5. I mean, do you prefer an article that says "Intel today introduced its new attempt to detrone AMD from its dominant position. However, the new i5 is a more expensive, slower processor that is waste of time testing at stock configuration (or overclocked to stable 24/7). Instead of wasting your time reading a pointless article, just go get yourself 965BE and live happily ever after."
If you ask Gary, I am sure he could e-mail you this massage every morning and you would not have to ever come to AT or any other site on Intel PR.
Just before you go all out with your AMD Fanboyism at me, all three of my computers have AMD CPUs. (I built two and the third is a laptop).
we also want to see temperatures when the processors are overclocked because some reviews say that when overcloded lynnfield reaches 95 degrees centigrades even when they were using a Thermalright MUX 120 which is ridiculous.
we also want to see the power consumption when the processors are overclocked because some reviews say that lynnfield power consumption skyrockets when overclocked.
I guess they should disable SpeedStep/PowerNow! in the mobile processor reviews too, you know, to test battery life at stock speeds without "cheating".
Oh snakeoil, I love you like a father must love his constantly underachieving, yet futilely continuing second born son. Always trying, never giving up and sure as bloody hell fighting till the last breath.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
159 Comments
Back to Article
moshpit - Tuesday, September 15, 2009 - link
When I read the part about somebody complaining about "illegal benchmarking methods", I had to sign up, JUST to complain about the person who would make such a silly comment. Then, I saw snakeoil's name involved in the comments section and him throwing around a bunch of FUD and using the word illegal down there as well.At that point 1+1 suddenly became clear and I realized that the plague of stupidity known as snakeoil is as much of a problem on this site as he is over on the techreport where I frequent more often.
Knowing that now, I want to apologize for any time the author of this update spent addressing that moonshot complaint as it's obvious you didn't know what you were dealing with. snakeoil is known on many forums for being a FUD spewing fanboy with delusions of grandure. Don't waste any more time worrying about comments about "illegal benchmarking procedures" when that little snot wouldn't know what the term "legality" even means.
Looking forward to further updates. And the numbers in this one are exactly as expected, no surprises for anybody with REALISTIC expectations.
snakeoil - Monday, September 14, 2009 - link
needs to much voltage and temperatures are almost 100 degrees centigrade.''Cons: 1) Lotsa of voltage increase needed to go past 3.6
2) no good coolers *yet* (I purchased coolermaster hyper 212 plus, better than stock but....)
Other Thoughts: Everytime i get a processor from newegg its like i'm destined to get the bottom of the barrel. after seeing initial reviews from various websites i was thinking 4ghz would be easy to obtain. Well i hit 3.6 ghz on stock voltage stable with prime 95, took 1.336 volts to get 3.8 ghz, and finally took me 1.375 volts to get to 4.0 ghz. Could run prime95 stable @ 4.0 ghz temps hitting mid 90's but i guess i'll back off 4.0 ghz til i find out more on this processor. initial reviews stated you wanted to stay under 1.4v, but i'm iffy on that if i want this thing to run for 6 months. kinda wish i woulda opted for the i920, i recommend anyone does if that can get a d0 stepping.''
http://www.newegg.com/Product/ProductReview.aspx?I...">http://www.newegg.com/Product/ProductReview.aspx?I...
Scalarscience - Friday, September 18, 2009 - link
Comparing a product just off the assembly line with a d0 stepping in an 11 month 'mature' process...? Of course the latter will typically have much more headroom due to its refinement (ie, what is your point?)As for newegg, they're the same as any other vendor. You either get retail or OEM, and I highly doubt Intel or AMD are sorting lesser binned parts specifically for them. The fact that newegg often has parts in advance of other retail outlets might suggest that you're often choosing earlier samplings from the production line though?
And I agree with the below poster, your crossover from TechReport.com isn't going unnoticed.
CB434 - Monday, September 14, 2009 - link
http://www.hardocp.com/article/2009/09/07/intel_ly...">http://www.hardocp.com/article/2009/09/...ntel_lyn...Konstantineb - Monday, September 14, 2009 - link
A full review of overclocking the lynnfield processors on stock voltages would be extremely useful.A couple of questions for the reviewer:
Is it possible to set the Cpu mulitplier of the Core i5 750 at higher than 20x, 22x for example?
How high can the memory multiplier be set with a Core I5 750 processor?
And lastly, in the previous review of the lynnfield processors you mentioned that the PCIe controller on Lynnfields can't be supplied by power seperately, and that you need to increase the processors
vcore on high clocks to achieve stable overclocking.
My queston is: If we can set the multiplier at 22x and and FSB at 160MHZ, "160 x 10 = 1600MHz frequency of the memory", aka rasing the multiplier and keeping the base clock as low as possible, will we be able to achieve stable stock voltage overclock?
A response would be greatly appreciated.
maxxcool - Monday, September 14, 2009 - link
I think you did a great job. i think the review is perfect in that you took the chip EXACTLY as a consumer would see and use it.That's it. and its fair.
I love it. Good job and good work. and Ban those fools who keep trolling on the issue.
Scalarscience - Friday, September 18, 2009 - link
I agree about the 'how an average end user would use it' perspective, and have posted such elsewhere (where we were discussing Lynnfield benchmarks and this site's were included.)Nomale - Monday, September 14, 2009 - link
I find it kind of important that Lynnfield seems to require added voltage to do even modest oc, since added voltage affects temperature and life expectancy of the CPU. From what I understand this is due to the on-die pcie controller which isn't separated from the rest of the CPU "voltage wise" and needs more more voltage when overclocking. Anyway, do you at present have any indications if/when this is going to change in future revisions?Alberto - Monday, September 14, 2009 - link
We have a test on a particular SW, still it's interesting.In a clock to clock comparison at 3.8Ghz thanks to a good cooling, K10.5 is 16% slower than Nehalem without HT, 34% slower than Nehalem with HT.
This meas that, under this SW, the Nehalem cpus are now pretty limited to stay in the power limits of 95 W/130W from the cpu power management. The 3.4Ghz Phenom has not problems with it's own 140W power budget.
The upcoming Intel 32nm cpu line looks a killer !!!!
oc3an - Sunday, September 13, 2009 - link
I noticed that the P55 chipset has an integrated Intel Ethernet MAC, yet all of the boards I have seen so far make use of a 3rd party (i.e. Broadcom or Realtek) chip.Does anyone know why this is?
HDThoreau - Sunday, September 13, 2009 - link
So can somebody guesstimate what an i7 920 overclocked to 4Ghz(45C) on air compares with among the i5's overclocked or not?The i7 920 is $200 at Microcenter. For the person which doesn't overclock anything ever, I can see how this article applies. But for people that build their own systems, I'm not seeing the big bonus here.
The comparable i5, which I haven't heard good things about it's 4 core overclock, is running at 20 bucks more. And the P55 motherboards ain't any cheaper than the x58 yet.
So, in summary. What is the advantage of buying an i5 anything rather than an i7 920 clocked at 4Ghz on air? That's it. Thanks.
Twoboxer - Sunday, September 13, 2009 - link
For those consumers that DO NOT overclock (and they buy MOST of the chips), the i5 is faster than the 965 out of the box.For those who DO overclock, two things are known:
- The i5 can OC to the same clock as the 965, and
- At the same clock, the i5 is faster than the 965.
WTF is all the arguing about?
Ann3x - Monday, September 14, 2009 - link
The i5 is faster than the 965 at equal clocks??? Lol is that a serious comment?You mean faster:
Except for the lack of HT
Except for the slower PCI-E
Except for the lower memory bandwidth
Except for the lower binning.
Yeeeaaahhhhh. In fact even if you ignore these 4 points theyre going to be the EXACT same speed. So youre wrong on multiple points.
The i5 is simply not faster than the 965. Now better value than the 965???? Hell yes.... Faster, just no.
Nich0 - Monday, September 14, 2009 - link
It sure is not clear, especially because we were talking about Intel processors, but I think that by 965 he meant the Phenom 2 variety. Which is why we got a very interesting -NOT- rebuttal from snakeoil.snakeoil - Sunday, September 13, 2009 - link
thats false.papapapapapapapababy - Sunday, September 13, 2009 - link
download this bench, and set everything to max (no med@ low res BS...)
http://www.nzone.com/object/nzone_re5_downloads.ht...">http://www.nzone.com/object/nzone_re5_downloads.ht...
my E6600 3.0Ghz awaits...
CB434 - Sunday, September 13, 2009 - link
Gary, you really know how to keep us on the edge our seats.deadrats - Saturday, September 12, 2009 - link
ok people, i have read most of the comments posted by both the obviously rabid intel fans and the equally rabid amd fans and there is one thing that absolutely must be acknowledged...intel's turbo feature IS overclocking, plain and simple.overclocking means to run a processor at a higher than rated speed by either raising the fsb (even with cpu's that have integrated memory controllers the bios still lists a fsb for simplicities sake) and/or increasing the default multiplier.
with the core i5/i7 we have just that: a dynamic increasing of the clock speed from the rated default speed and said increase is only activated when the cpu is under load.
the core i5 750 has a rated clock speed of 2.66 ghz but the fact of the matter is that under most conditions it will be running at 3.2 ghz, it's absurd to compare it to a core 2 or phenom 2 running in the 2.66 ghz range and proclaim the i5 750 as the superior chip when the 750 is benefiting from 600 mhz clock speed advantage.
it reminds me of the muscle cars of the 60's when car companies would purposely under state the horsepower ratings by measuring horsepower and torque at a point below the rpm range when peak power was made, so that it could look like their cars needed less horsepower to be faster than the competition.
in fact, as late as circa 2000 chevy released a camaro ss that was rated at 305 horsepower that was faster than the corvettes of the day, hell the damn thing was ripping a 13 sec flat 1/4 mile time and anyone that ever drove one could tell you that it made way more than 305 hp. one magazine did put it on a dyno and found that it made about 350 hp at the wheels which works out to about 380 or so at the fly wheel.
same thing with intel: they market a chip as being a 2.66 ghz part and then under any condition than idle have it run at 3-3.2 ghz, my question is why not just clock it by default at 3.2 ghz, sell it as such and simple allow it to be down clocked when idle to save power, just the way they have done with the penryn and conroe and the way amd does with it's processors.
turbo mode is a marketing gimmick, nothing more and if i did buy a i5 750 (and i'm seriously considering picking one up) i would turn turbo off, buy a good after market cooler and overclock the chip myself to about the 3.5 ghz range, which should prove to be a very stable over clock.
HavocX - Monday, September 14, 2009 - link
"overclocking means to run a processor at a higher than rated speed""with the core i5/i7 we have just that: a dynamic increasing of the clock speed from the rated default speed and said increase is only activated when the cpu is under load."
Higher than the default speed, yes. Higher than the rated speed, no.
Overclocking means running the chip at a higher speed than the manufacturers settings. Variable clock speed is not overclocking. A chip at stock setting are never overclocked.
That said testing maximum overclock at similar cooling situations is also interesting. What is never interesting is comparing chips at worse than their out of the box settings.
"the core i5 750 has a rated clock speed of 2.66 ghz but the fact of the matter is that under most conditions it will be running at 3.2 ghz, it's absurd to compare it to a core 2 or phenom 2 running in the 2.66 ghz range and proclaim the i5 750 as the superior chip when the 750 is benefiting from 600 mhz clock speed advantage."
This is true, it should be compared to chips of the same price and power consumption. And this comparison is generally favorable to the i5.
"they market a chip as being a 2.66 ghz part and then under any condition than idle have it run at 3-3.2 ghz, my question is why not just clock it by default at 3.2 ghz,"
Because that would allow the chip to run at these speeds at full load on all cores, and they do not want the power/heat to ever raise to those levels. The beauty of turbo mode is that you maxize the performance withing the set power/heat limit. No competing chip do that, overclocked or not.
TA152H - Sunday, September 13, 2009 - link
What's Montague? It is nor hand, nor foot,Nor arm, nor face, nor any other part
Belonging to a man. O, be some other name!
What's in a name? That which we call a rose
By any other name would smell as sweet.
So Romeo would, were he not Romeo called,
Retain that dear perfection which he owes
Without that title. Romeo, doff thy name,
And for that name which is no part of thee
Take all myself.
This is for all you idiots that keep talking about the term overclocking, like it's a meaningful conversation.
I'll agree turbo mode is a stupid name, since you're not using forced induction, but the technology itself is valid. Idiots can argue over the term, but it gives one a finer granularity on increasing clock speed than a simple overclock would. It's intelligent in the sense it overclocks based on load, and it's a perfectly valid and useful feature for the vast majority of people buying the Lynnfields. They aren't going to overclock their processors in the conventional sense, and get an out of the box experience that gives them extra performance, without violating thermal and electrical characteristics of their processor and system. It's a good technology for a lot of people. Call it whatever you want. That's not the important aspect of it.
I agree with most people here that it's pretty useless, since most of us would just get an i7 920, or, for those who don't want to spend so much, get a i5 750 and overclock it. But, most people buying computers don't do this, and Intel makes most of their money on these people. It's a really good technology for them.
Ann3x - Sunday, September 13, 2009 - link
Totally agree its great for mass market level (turbo mode).Virtually noone who buys a dell desktop is gonna want (or know how) to overclock.
I simply disagree that this makes 90% of the i7 line redundant, especially when youre talking to a tech audience. To an enthusiast (and most people whore going to buy x58 are going to be this level) the 920 is still imo, the best buy *in most cases*.
TA152H - Monday, September 14, 2009 - link
Ann,I've been saying that all along.
I completely agree. I'd much rather have the i7 920 than the brain-damaged Lynnfield platform. But, let's keep it in perspective. Say I don't have that much money to spend, and my choice isn't an i7 920, then the i5 750 comes into play. Only when the Lynnfield isn't competing with the Bloomfield does it make sense for the overclocking crowd. But, there are some price points the Bloomfield can't reach, but the Lynnfield can. At that point, you have to compare it against equal cost platforms, and in that context it could be an attractive processor even for people that are comfortable overclocking.
But, again, I am one of the biggest proponents of the i7 920 on these forums, so much so that people twist my words on it. Outside of my instinctive dislike of absolutes, I would say the 870 and 860 are essentially worthless processors for the technical savvy, as is the i7 950 (I still can't bring myself to putting the i7 before the brain-damaged Lynnfields). But, as easy as it is to say you're better off spending the extra $100 or $150 for the i7 920 rather than getting saddled with a castrated platform, that's an entirely specious argument if the person buying it simply can't pay it.
I would say the biggest problem with the i5 750 is the lack of IGP to go with it. It's a Celeron, without the platform to go with it. In the broader market, this might prove to be a bigger impediment to widespread adoption than the emasculation the chip suffered from. While you can add a discrete card, it adds cost, chews up a slot, and probably generates more heat. It's not horrible, but not ideal for the target audience.
Inkie - Monday, September 14, 2009 - link
"I would say the biggest problem with the i5 750 is the lack of IGP to go with it. It's a Celeron, without the platform to go with it."If you believe that a $200 processor is a Celeron then you seem to have little knowledge of current market conditions. That's without considering the excellent performance of Ci5. Clarkdale is what you are looking for, not that anyone will call Clarkdale a Celeron, except perhaps those that look down their noses on dual-core.
Ann3x - Monday, September 14, 2009 - link
Im not disagreeing with you in anyway way ^^. Totally on the same page here.Lynfields great.
Just not the revolution the reviews are making out that it is.
CB434 - Sunday, September 13, 2009 - link
"the core i5 750 has a rated clock speed of 2.66 ghz"The i5's rated speed is 2.66 stock. 2.93 4 cores active, 2.93 with 3 active, 3.06 with 2 active and 3.2 with 1 active. That's what it runs at "stock" with default BIOS settings. I don't know (or care) whether it's OC'ing or not, I just know that it runs at those speeds by default. And it does so in a very effecient manner.
The turbo isn't a gimmick. If it was I wouldn't be using it.
I will be buying i5 and Noctua cooler. If I was buying i7 or Phenom II (which was a big possibility) I would be overclocking it to 3.6-3.8 24/7 stability permitting.
With i5 turbo on, I can overclock to 3.4, consuming less power and still reach 4Ghz in single core apps. While running COOLER, and less VOLTAGE then the Phenom and i7 alternatives at this same speed. It handles all of this on the fly. Without needing different OC profiles for different games or apps. Do you know that when only only 1 or 2 cores are active the heat and voltage necessary is alot less? That's why the single core apps can run at 4 Ghz safely. Unlike the Phenom or i7, which can only get to 4Ghz while OC'ing all cores instead of 1. It's the same reason why you can OC a Phenom II 550BE higher then a 955BE. And why when you unlock the extra two cores on the 550BE, you have to reduce your OC.
This isn't a gimmick this is effeciency at it's best. I'm sure in the future there will be better solutions, but for now I am delighted with this. It's a powerful single core, duel core and quad core all in one. It's like a 550BE + a 955BE. You get the strengths of all of them with none the weaknesses. Since I am primarily interested in gaming, the lack of HT isn't an issue. And the AMDs don't even have HT anyway. With this, "turning turbo off to OC" becomes a thing of the past. Unless you like ineffeciency.
This i5 will be my first Intel CPU since Celeron 766mhz.
Eeqmcsq - Sunday, September 13, 2009 - link
"It's a powerful single core, duel core and quad core all in one."That's a great way to describe Turbo mode. It ends the question of whether to get a faster dual core or a slower quad core, since you can get both. I suspect you can also force the CPU to stay as a faster dual core by disabling 2 cores at the OS level.
snakeoil - Sunday, September 13, 2009 - link
No,turbo is just overclocking.I'm going to explain you the real story behind this intel turbo mode.
intel was having a problem with core i7 920, this processor was the only affordable core i7 but it was killing the more expensive core i7 (950,975) because can be overclocked to 4.0 ghz and obtain the same performance for less money even when it has a locked multiplier.
we need to kill overclocking before overclocking kills our profits intel thought.
So intel decided to create the turbo crap story, or auto overclock.
Instead of giving overclocking as a free bonus we will create this turbo feature and begin charging for it.
that's true intel is now charging their users for the overclocking which with amd phenom 2 is free.
So intel is using all the available overclocking the processor may have and call it turbo (like turbo time in toy story).
This way they can segment their lineup, how? because they make a model overclock 600 mhz, and they can make the turbo more or less aggressive depending on their marketing needs.
Turbo mode is just overclocking, is not fairy dust, is hocus pocus
intel uses to skin some intel ignorant users.
Intel never liked the idea of giving overclocking for free, now has decided to kill overclocking because turbo auto auto overclocking consumes all available overclocking headroom, and intel can use it to charge more for a processor and to segment properly their lineup.
what's wrong with intel?
Nich0 - Sunday, September 13, 2009 - link
In the end it doesn't matter if Turbo is overclocking or not. This is a silly argument about semantics. What matters is that in their stock configuration it's fair to compare the Lynnfield with other CPUs also in their stock configuration. Which means that it's reasonable to compare an i5 750 (stock, i.e. Turbo is left on) with an equivalent AMD processor. Question is: what is an equivalent AMD processor? DIfficult to say, really.Where I live, for example, the 95W version of the 3GHz 945 is priced similarly to the 750 (it's actually a bit more expensive, but I'm guessing that this might be different in other countries), so that would be a good candidate. Similar price, similar thermal enveloppe, similar frequencies (remember that Turbo will push the frequency of the 750 between 2.8 and 3.2GHz).
But I am tempted to say: who cares, apart from the fanboys?
Ann3x - Sunday, September 13, 2009 - link
It does matter if reviewers are going to completely dismiss CPUs with no or less aggressive turbo modes.The inital review said that there was no point in any i7 below the 955. This is clearly an absurd assertion.
It is basically just because of the higher turbo mode enabled results taht anand has made this statement. As weve discussed turbo mode is really manufacturer endorsed overclocking (with slightly more sophistication) so surely the review is flawed as youre not comparing like for like.
Anyway....
Nich0 - Sunday, September 13, 2009 - link
I take it you're one of the second group of people who take exceptions with the article: no not the AMD fanboys but the proud owners of a Bloomfield, probably a 920 - or about to get one? What's wrong? Overnight it didn't become a bad CPU. It's still a fantastic piece of silicon, I'd say. Anyone who says the contrary is deluded (snakeoil anyone?). What one need to understand is that Anand's statement implies something along the lines of: with all things being equal (like retail channels maturity, which at the moment gives Bloomfield an edge) and for prospective buyers that don't need the 32 lanes of PCIE, etc...Why get so upset?
So-called 'manufacturer endorsed overclocking' has very little relevance. It's semantics. AMD is doing exactly that with their Phenom2s. Ridiculous voltages that pushes the thermal enveloppe. So what? There are 2 states for a processor: its drop-in stock state and its altered (either manually or via an auto mobo thing), i.e. OC'd state. What's wrong with comparing CPUs in their stock config? WHat's wrong with comparing OC'd CPUs at similar frequencies? I don't see it.
Ann3x - Sunday, September 13, 2009 - link
Oh so were going for the "you must be an XXXX" tactic..... Implied bias if you cant actually argue the point. Ok then.No, actually I run multiple PCs, a Q6600 a 920, a C2D and a phenom 2 for a variety of purposes (server, gaming x2, work and multimedia), I also do some system building (and pre sale overclocking). I dont intend to change machines personally in the next few months so consider me an interested observer.
What I object to is simply the supposedly unbiased reviewer making such a retarded statement.
I totally agree that the new chips are pretty good. However the review makes it sound like theyre a huge leap forward when infact theyre simply intel realising they have headroom in their architecture.
Hell, Im much more interested in the integrated PCI-E than the gimmick of turbo mode.
As for the semantics thing. Yes it is semantics. Hence why Im bothered about anand making a such a song and dance about it. The new i7s are damn good cpus but to suggest the performance increase is due to anything other than the "overclock" of the aggressive turbo mode (which is something attainable EASILY by the old i7s) is totally fanciful.
Inkie - Sunday, September 13, 2009 - link
Turbo Mode is not a gimmick. It offers real world benefits in real world situations, without exceeding the specifications of the processor.Ann3x - Monday, September 14, 2009 - link
Look, I totally agree an aggressive turbo mode is a good thing for the average consumer. The merits of turbo mode are not in debate. Its unequivocally a good thing for the average buyer.The ONLY thing I disagree with about the review is the massive amount of guff thats been put out making out that turbo mode is a revolution. Its really not.
Its quite simply intel realising the headroom they have in their architecture and cleverly doing something useful with it (coincidentally exactly what overclocking also does).
The assessment of this is what I disagree with.
When a review site is aiming at a tech audience (like anand does) surely there has to be some differentiation between the audience and the average consumer? If the review had said something like "For the non technical user the new CPUs offer much more than the current i7 lineup" Then Idve totally agreed with it. As it was the review uses the emotive and totally wrong assertation that "there no point in any i7 below the 965". Thats just stupid.
Imo actually the real situation at enthusiast level is that there is NO justification (unless money is genuinely no object) in any i7 except the new CPUs (for people who dont know how/are unwilling to overclock) and the 920 (as it pretty much always clocks to similar levels as the 965 anyway).
So, to me, anand totally missed the point (given the audience) with the inital assessment. This is what disappointed me, this is what Im bothering to argue about.
And for the record its highly ironic that the people calling others "fanboys and morons" tend to be the ones who dont back up theri points. Take a step back, THINK and perhaps youll see that some of us are actually talking sense rather than blindly reading a review and cheer leading.
Inkie - Monday, September 14, 2009 - link
Not quite sure why that was a reply to me. Apparently you agree with what I said. You called Turbo Mode a gimmick. Now you say it is not a gimmick. End of story.Ann3x - Monday, September 14, 2009 - link
It is a gimmick in that it is just what we had before (overclocking) wrapped up in marketing speak. Nothing new just assistance for the brain dead.It is however a useful gimmick as the majority of consumers are retarded and dont know how to get the most from their CPUs.
Clear enough for you smartass?
Inkie - Tuesday, September 15, 2009 - link
So, before we could selectively increase the clock on particularly cores according to core loading conditions? Perhaps looking at the Xeons will help you understand. One model is a 45W TDP model with a 1.86GHz clock. However, it has the ability to Turbo past 3GHz, all within the specifications of the processor.Nich0 - Monday, September 14, 2009 - link
I don't know if your last paragraph was aimed at me, but I didn't call YOU a fanboy. OK maybe categorizing you as a 920 owner might have been a 'cheap shot' but it sure did hit the target.Now to come back to matters at hand. Nobody denies that the 920, 750 and 860/870 are based on the same micro-architecture. And Gary's preliminary results of these CPU OC'd at the same frequencies show near identical performance. But the main difference is that now we have a mainstream processor as opposed to an enthusiast, which means that in a lot of cases it's going to be left in its stock state: large OEMs like Dell or HP, people who can't or won't OC, etc... And for these situations, it seems that for example, the 860 is a better processor than the 920. Sure it's a close match but at least the 920 keeps its dignity. Imagine if Intel had come out with the 860 at 3.06GHz with no Turbo in 3/4C and just a 3 Turbo in 1/2C. It would have been a massive slap in the 920 owners face. Also the Turbo allows Intel to keep the TDP to reasonable levels. The 860 is probably more often than not a 3.06GHz (minimum) processor but you can't call it that because it will throttle back to 2.8GHz in some circumstances. And it's OC'd performance is almost identical to that of the 920 (it would appear). Again there are some cases where you are better served by the 920. Clearly. But the fact remains that soon enough you will have combos mobo+860 a good deal cheaper than 920+mobo and for 'better' performance. Hence Anand's statement.
Roland00 - Saturday, September 12, 2009 - link
this is exactly what i was looking for, I am waiting for the other tests you will post later.I sell prebuilt oem computers, hp/compaq, gateway/acer, lenovo, etc so the first article is what I need to know for my business.
At the same time, I build my computer as well as several of my friends so what a cpu oc to matters a lot and what I recommend on that front.
setzer - Saturday, September 12, 2009 - link
I know this might come too late for this specific review, but maybe you guys could do this one later.Okay, most of the reviews seen for the i5/i7 cpus are compared to amd's top of the line Phenom II X4 965 BE, while this is perfectly valid to compare top of the line cpu's, it's also a 140W processor, and the power consumption tests show that, also the board used to test said cpus is also a top of the line board which in my opinion doesn't compare well with P55 based boards and their moto.
So what I'm asking is if it's possible to compare offers from AMD and Intel in their usual configurations as in, mid-range.
This would probably be a 785G board paired with a 65w (P2 x4 905e/x3 705e) or 95w processor (P2 x4 945/P2 x4 720).
For the intel camp, this would be a P45 board with the usual suspects (up to the Q9550 and their low power versions if you have them around).
As for the new brothers in arms, the i5 750 and i7 860 would be my choices (even if the 860 has a current asking price a bit out of the league of the previous cpus, it's the only new one with ht enabled).
Don't take me wrong, I, like most of the readers, like to see new hardware pushed to the limit, but at the end of the day my personal computing goals are much more mid-range oriented and I just don't see justification in passing the monetary range of the choices I listed above.
Think of it as AT for the masses of poor economy people that still want something new :P
coconutboy - Saturday, September 12, 2009 - link
Saw the drama queens throwing their hissy fit over how things "should" be tested. Those clowns need to realize they do NOT speak for most users, they're just a vocal minority. I skipped most of the remaining comments so sorry if I missed something.I'm building two new systems, one w/ a i7 920 and another with a either an i5 750 or i7 860. My lady and I want these specifically for gaming and plan on doing mild OCing w/ minimal voltage increase (preferably none, but looks like p55-based cpus demand at least a tad extra voltage). Gaming benchmarks showing this scenario are helpful and I look forward to updates/articles comparing p55 versus x58.
More wish list: with the lowered power draw of p55 setups, I'd love to see what's being sucked outta our wall outlets under OC'd/SLI/Xfire setups since this can mean an extra ~$50 for a beefier power supply.
Pneumothorax - Saturday, September 12, 2009 - link
Anyone else notice the extra 0.1V required to o/c the 1156 cpus vs the 920?Nich0 - Saturday, September 12, 2009 - link
It might have something to do with the PCIE. As soon as you raise the BCLK, it overclocks the PCIE too which requires some voltage to stabilize. More than the actual processing cores it seems.CB434 - Friday, September 11, 2009 - link
Very good work Gary, for all of the articles.Clock for clock the i5 is faster. That is why I chose it before i5 was even released. The only advantge Phenom II has is 16x CF (but no SLI). No price advantage either.
If you OC both to 3.6 or 3.8 Ghz, i5 wins. Unless it's a game/program that favours the AMDs in which case, the Phenom will beat the i7 as well.
Stock clocks are most favourable for Phenom since it's 3.2Ghz vs 2.66Ghz. But that's irrelevant for me, because whether I get Phenom/i5/i7 I would OC to 3.6/3.8 with Noctua cooler. When they are both are 3.6Ghz, not only does the i5 consume less power but it's also faster, especially in CPU orientated tasks.
Since turbo is so effecient, I plan to OC to 3.5Ghz with turbo on, and that will give 4Ghz in single core apps, 3.9 in dual core apps etc. While saving power at the same time. Fantastic!
It's a no brainer choice for me, and I'm not an Intel fanboy (I use Athlon currently), I just want best bang for buck.
I'm looking foward to the followup when it is posted to confirm my thoughts.
tajmahal - Friday, September 11, 2009 - link
Actually you can get a 780a or 790a Nvidia motherboard for 16x-16x SLI. Since the PII 955 BE can also hit 3.8GHz when overclocked, there's no reason to pay the extra $50 for the 965.CB434 - Friday, September 11, 2009 - link
Agree about the 955. At the shop I will be buying from online in my country (cheapest in Australia) i5 is $285, 955 is $271. $14 difference. Gigabyte p55 UDP5 board is same price as the 790FX AMD board. For me the extra performance and effeciency (heat/power) is worth the extra few pennies.Unfortunately regarding the AMD SLI mobos's, this shop doesn't carry them. But even if they did, it's the same problem. SLI only. It's either CF only or SLI only. If you are going to go the AMD way, your probably better off with the full Dragon platform with the AMD software to take advantage of it.
Dudlington - Friday, September 11, 2009 - link
What I haven't seen any reviews talk about is this:What if I want to have 2 x Video cards, along with another expansion card that uses PCIe 1x or 4x? Is it just not possible with P55?
MadMan007 - Friday, September 11, 2009 - link
Agreed, although dual graphoics cards don't interest me personally knowing how moterboads assign PCIe lanes depending upon slot population is important. If you have two 1x devices and one happens to be in the secondary graphics slot, does the primary graphics slot get cut down to 8x? It probably wouldn't be a huge performance hit but it's nice to know this type of detailed information.Stradigos - Friday, September 11, 2009 - link
So when's that mobo round up gonna happen? I've been refreshing the page all day! I thought you said Thursday/Friday we were going to see the first article, and then Monday the last? No rush... just can't wait to read it! :DAffectionate-Bed-980 - Monday, September 14, 2009 - link
What about the mobo roundup from X58? We got like 3-4 mobos in a "roundup" when Toms used 3-4 SERIES of roundups each with 8 mobos (over 2 dozen mobos total).What about that i7 overclocking guide that never came? What about SSD stuff? Huh? The DDR3 roundup?
If you go back article by article a lot of them promise MORE content "later this week" or whatever. Nothing EVER comes. I wouldn't get my hopes up about AT articles anymore. They're quality when delivered but if it's a "preview" with "more to come this week" or a "lab update," I wouldn't expect much at all.
dragunover - Friday, September 11, 2009 - link
Is the clock for clock comparison, with an average overclock. This represents a huge deal for me in comparing processors' actual raw speed compared to just wildly varying clock levels.crimson117 - Friday, September 11, 2009 - link
I'm torn... AMD 965 BE for the most part beats Intel i5 750 non-turbo, but costs $50 more.But with turbo, they're about even.
But turbo is just single-core on-demand overclocking, right? Doesn't AMD have some utility (Overdrive or Fusion?) that automatically OC's a bit?
It think that'd be a more fair comparison: Auto-mobo-OC'd AMD 965 BE vs Turbo-enabled Intel i5 750.
MadMan007 - Friday, September 11, 2009 - link
As part of your decision process I think you should note that this particular test, whlie no doubt one of just many, was of a pretty well threaded application. Turboboost *still* provides a notable benefit in even well threaded apps. Turboboost is more complex than you implied, just check the main review it's more than single-core oc'ing on demand.Nich0 - Friday, September 11, 2009 - link
What's fair and what's not leaves a lot to interpretation it seems. To me a fair comparison would be:1- 965 stock vs 750 stock (ie Turbo is on because that's the way the CPU is regardless of the motherboard). People who won't/can't change BIOS settings. Let's say they bought a DELL, for example.
2- 965 3.8GHz vs 750 3.8GHz (Turbo off to get there). Gary did it. Check.
3- 965 Auto OC vs 750 Auto OC (Turbo off). WHy not.
4- 965 Best OC vs 750 Best OC (Both on air, closed box). Similar to 2-, wouldn't you say?
The point is, there are a lot of ways to do these comparisons, and Gary is doing a pretty decent job to cover most of the bases IMHO.
jonup - Friday, September 11, 2009 - link
What happens if you are running a nVidia mobo?See, Turbo mode is not like a regular oc. It does not affect other component, power consumption, heat, etc. It really does not have the drawbacks of oc. I thank Anand for explaining it. If you want to do an OC some of the P55 boards come with utilities that will oc the system by a small percentage at windows start up. This will be equivalent to the AOD.
snakeoil - Friday, September 11, 2009 - link
no, turbo is just overclocking.Stradigos - Friday, September 11, 2009 - link
You are seriously retarded. Please go play in traffic. Turbo is not overclocking, and your ignorance is getting really tiring on this matter."Overclocking is the process of running a computer component at a higher clock rate (more clock cycles per second) than it was designed for or was specified by the manufacturer"
http://en.wikipedia.org/wiki/Overclock">http://en.wikipedia.org/wiki/Overclock
Did you completely miss the part where he explained why single cores clock higher than quad cores? It has to do with the amount of voltage that can be used. If you have a process that isn't using all four cores, why use them? So shut them off and pump the voltage through to the other cores. This is not overclocking. This is meant to happen by the designers, creators, and manufacturers.
MadMan007 - Friday, September 11, 2009 - link
As a corollary to your last paragraph I think we can conclude that snakeoil's brain, in addition to being a low neuron wet state device in the first place, permanently has turboboost turned off as well so most of his brain is permanently shut off and he can never perform mental processes well.coolcatmatt - Saturday, September 12, 2009 - link
What is turbo boost? Intel has answered that question for us.http://www.intel.com/technology/turboboost/">http://www.intel.com/technology/turboboost/
"It automatically allows processor cores to run faster than the base operating frequency if it's operating below power, current, and temperature specification limits."
Looks like overclocking to me too.
Ann3x - Saturday, September 12, 2009 - link
No.....Intels marketing department says its not overclocking (yay) so people who actually believe everything is true in a marketing speil get suckered into it.
Turbo mode is just manufacturer stamped overclocking with some bells and whistles. Snakeoil is actually right but guess its easier to flame him because of the other things hes written.
So much missinformation around on overclocking (increase thermal envelope, increase power consumption etc etc) most of the points only apply IF you overvolt. Which you really dont need to do on most i7s.
But hey overclockings bad mkay (unless its written in a fancy marketing name on the side of the box, in which case its revolutionary).
Inkie - Sunday, September 13, 2009 - link
That would be a neat trick: increasing clockspeed on all 4 cores, thereby doing more work per unit time, yet consuming no more power because you didn't overvolt? So much misinformation around on overclocking.....Ann3x - Sunday, September 13, 2009 - link
Not no more power, negligibly more power.Anyway does the turbo mode not increase clock speed? Oh wow. Guess they have some way to beat P = C * F * V^2?
Inkie - Sunday, September 13, 2009 - link
I suppose it's what you mean by 'negligible' then. That carries a diferent connotation for processors than it does nuclear power stations.Yes, turbo mode increases clock speed on the cores in question. What is the point you want to make? It is all done within the specifications of the processor. It's just a clever application of the idea that if all the cores are not working hard then there is room to clock some of the cores higher. If you want to call that clever overclocking then you can, but what most people mean by overclocking is exceeding the specifications of the processor.
Alkapwn - Friday, September 11, 2009 - link
Begone, foul demon of misinformation and bias, begone.Alkapwn - Friday, September 11, 2009 - link
The prior was @ snakeoil.jonup - Friday, September 11, 2009 - link
A dynamic OC which does not change the thermal envelop. 965BE has 140W TDP at stock. When you overclock the the thermal envelop skyrockets, putting extraordinary presure on the power circatery. Turbo does not do that. Yes, it increases the clock of the individual cores, but does not put additional pressure on the system, it is the default setting, it is not user-initiated. It is the stock clock. It just varies. The clock is not set in stone. If I go and change internal clock then it is OC. It is over the clock it was meant to be.avaughan - Friday, September 11, 2009 - link
Jon Stokes at Ars Technica noticed something odd in the last graph at http://techreport.com/articles.x/17545/6">http://techreport.com/articles.x/17545/6 . Basically Far Cry 2 at 1600x1200 4xAA is GPU limited (expected), but the Phenon II x4s and Core 2 Quads all converge on a frame rate that's approx 10 fps higher than the i5-750/i7-870 (unexpected).The E8600 even beats the i7-975 at 1600x1200 4xAA, even though Far Cry 2 is obviously cpu limited at lower resolutions.
Any chance you guys could try to reproduce/investigate that, and look at single gpu resolution scaling to at leats 1920x1200.
I know that's the sort of thing you mainly look at with new chipsets, but with the northbridge and cpu merging i5/i7-8xx is effectively a new chipset.
Gary Key - Friday, September 11, 2009 - link
I have spent the better part of the last twenty four hours trying to figure out our FarCry 2 numbers and it also happens in other games by the way. ;) The only conclusion I have right now is that with the latest NV drivers, Win7, and certain games, the Phenom II 965 is faster now.Not only against Lynnfield but also against Bloomfield in the tests we utilize for the motherboards. The CPU oriented action test we use in the processor benchmarks tells another story. So something is happening along the graphics path at this point, what exactly is something we are trying to isolate.
avaughan - Friday, September 11, 2009 - link
Ah thanks. Good to know that you're looking into it. Does it seem to happen in many games, and is it limited to NVidia cards + win 7 only?Since you can reproduce it, it's probably a good idea to make a note/show some benchmarks demonstrating it in an article soon, before too many people start replacing their existing Phenom II/Core 2 Quads looking for better framerates at high resolutions.
Hlafordlaes - Friday, September 11, 2009 - link
After reading this update and the comments, I went back to peruse the comments on the launch article. Ugly!Granted, there is something to be said for improving comparisons at equal clocks, etc., but I can't help but get the feeling we are ignoring the writing on the wall. In short, the desktop PC (not servers or workstations) is rapidly going the way of the dodo. The last remaining performance-demanding app on the PC is gaming, and PC gaming is dying in its present form.
So the home PC trend is toward a console/nettop in the living room and maybe a NAS for local storage, synch services for phones and note/netbooks, tied to cloud backup. The P55 platform is a clear step in the direction of SOC, enabling Intel to stay relevant eventually in the living room (notice the high % of uATX boards, too). When and if Larrabee comes into play as a competitive graphics player, we might see an eventual renaissance of PC gaming, but on what has virtually become an Intel console, not a desktop. This is why P55 makes sense to me; it forks the market so that X58 and so on forms the backbone of servers and workstations, and P55 drives in the direction of consoles.
Unfortunately for NVidia, I do not see a space for them in the midterm, barring some miracle alignment of the stars.
darkos - Friday, September 11, 2009 - link
Microsoft Flight Simulator is a great tool for testing these multi-core setups. The latest service pack includes the ability to use multiple threads, and has support for both DX9 and DX10.There are many flight sim people out here that want to know how these systems will fare with FSX and x-plane.
Gary Key - Friday, September 11, 2009 - link
I actually have an excellent FSX benchmark. I did not run that app for the mobo reviews but let me see what I can in a blog or something next week.darkos - Monday, September 21, 2009 - link
Any word on those FSX benchmarks?darkos - Friday, September 11, 2009 - link
that would be great! thank you very much. :)Nohliife - Friday, September 11, 2009 - link
In your review the multi-gpu sli and especially crossfire didn't work well at all with Lynnfield. However over at Tom's hardware it didn't seem to work all that bad. Do you have any idea why?Is it the on-die pcie controller which causes problems or the shared x16 line? I would expect the x16 to be a bottleneck at very high resolutions but are there other scenarios it will cause problems?
ClownPuncher - Friday, September 11, 2009 - link
Keep in mind, 2 card sli and Xfire scale fine on 2 x8 lanes. Three and 4 card setups are where the PCIe bandwidth is not going to be up to snuff. If I recall... the tests were shown with 2 4870x2 cards, with a setup like that you would need full x16 lanes, therefore a LGA 1366 setup would be needed.Matty R - Friday, September 11, 2009 - link
Does overclocking disable the turbo mode even if there's some headroom left?Matty R - Friday, September 11, 2009 - link
So hypothetically speaking, I could clock it to 3.4 and when turbo kicked in it would throttle a couple of steps up to 4.0, maybe. That would be sweet.Nich0 - Saturday, September 12, 2009 - link
Preliminary results from Anand shows a best-case OC (with Turbo on) for the 750 of 3.2GHz and with Turbo it would reach 3.84GHz on a 2C/1C stress (although Anand shows 4.16GHz for that last number. I'm still waiting to hear from him about it). Go check the main P55 article, in the overclocking section.Nich0 - Friday, September 11, 2009 - link
It doesn't look like it does, so you'd better make sure there is a good amount of headroom if you leave it on.faxon - Friday, September 11, 2009 - link
i see it's listed on MSI's website now, and people have removed the cooler to discover a hydra chip. any word on if this will be benchmarked in the mobo lineup, or is it reserved for its own multigup lineup instead? and are there any other motherboards known to have this chip? really need to know ASAP, and i need multigpu scaling with NF200 as well, cause unless it's good enough (comparable to x58 performance with 5870x2+5870 or equiv usage of bandwidth), i need to get my hands on an D0 920 before intel stops making them so i can be ready when i get my 5870 to replace my UD3P lolfaxon - Friday, September 11, 2009 - link
i should also mention, this is with the 860 costing $30 more than the 920 at microcenter right now, and the fact that the S1156 boards with the PCIe lanes i need being at least $180-200 out the door, and ram being negligible since i would be selling another system when i get this one anyway. the 920 would be easier to overclock using less voltage, i dont care about the PCIe latency reduction since i would be using a bridge chip anyway, and im a RAM whore and an overclocker (4.2ghz quad and rising) so the S1366 is more ideal anyway in most ways. im just really eager to see if the hydra is all that it's cracked up to be, and i wanna see some OC results on the P55 FTW vs the EVGA X58 micro and standard (as far as ease of OC and max OC goes).Nich0 - Friday, September 11, 2009 - link
At the moment it looks like the 920 is a bit cheaper than the 860 so I think it's normal that people should question the wisdom of i7 1156 as opposed to i7 1366. But the 920 has been around a while so the retail channel has matured (lower prices, availability, stepping). When the 860 has reached the same level of maturity (if it does), I'd be surprised if the 860 is not cheaper than the 920. Anyway that could be a moot point because by then socket 1366 may have gone 6-core.strikeback03 - Friday, September 11, 2009 - link
Thanks for the additional tests. I didn't bother posting in the comments of the launch article as they were already extremely long, but I am also one of those looking for clock-for-clock comparisons. As the application I would be designing for is Photoshop, I'd want HT. This puts the i5 out of consideration for me, and as the platform cost of an i7 860 isn't a lot different than a i7 920, I approach it more along the lines of "Prove why I shouldn't just go straight to the 920 and keep the door open for future higher-performance options". For either one I would apply some easy overclock ( I was thinking more like 3.3GHz) and would be happier if the motherboard could do this automatically. I'm basically coming from more of a computer-as-a-tool than computer-as-a-hobby perspective, so I look forward to the new results.Scali - Friday, September 11, 2009 - link
Since my previous message may have gotten lost in the flood of flamewars, I'll just briefly state it here again, seems appropriate.I'd like you guys to also spend some time on reviewing the onboard audio features, particularly since there seems to be a new generation of VIA onboard chips. It's hard to get any info on what these are really capable of, and how mature the software/drivers are that ship with them. It could mean the difference between having to get an addon soundcard or not, when buying a motherboard.
Gary Key - Friday, September 11, 2009 - link
I will have a short synopsis on the audio choices provided on the P55 boards in the motherboard articles. Based on the updated VIA chipsets and Windows 7, we are thinking right now that a separate piece on the state of on-board audio would be wise. I will see what we can do to get that done this month. In the meantime, I like the VIA 2020 better than the Realtek ALC 889a from an overall perspective, the drivers still need a little work but audio quality is better in my opinion.Scali - Saturday, September 12, 2009 - link
That would be nice, Gary.One feature in particular that has made a big difference for me, was the support for realtime Dolby Digital Live encoding. When Realtek started offering this feature on their onboard chips, it meant that you could get digital 5.1 output to any Dolby-capable amp/home stereo. You were no longer depending on the analog outputs, which required lots of wires and generally was of poor quality with onboard solutions.
Later I got a Soundmax, which could do DTS Live encoding. The sound quality was excellent.
(I'm not talking about having stereo sound upconverted to 5.1, but about having 5.1 audio from games and such encoded in Dolby or DTS in realtime, and sent over the digital output).
So I'm particularly interested in what the VIA chips can do in that respect.
MadMan007 - Friday, September 11, 2009 - link
Just remember that there's a difference between judging chipsets and implementatins of said chipset. Especially for analog use there will be variation among implementations.yyrkoon - Friday, September 11, 2009 - link
Now comes the fine grain; This is where subjective opinion comes into play. What is "reliable" ? Apparently Subjective. What is too much attention on a per name brand basis ? Apparently Subjective.Let us toss out a fictional example.
Lets say product name A get a lot of attention, and has a strong "fan base" Product A is typically known as being the fastest on a per product test basis, but in the IT sector is not known as reliable. Product A gets tons of raving reviews despite this, even despite the fact that from time to time they also have some very serious problems. Product A does very well in the OEM sector ( probably with tons of OEM support from them working together as partners or not ).
On the flip side of things products name B has been in the business for an equal amount of time, or possibly longer. Product B is known mostly as a very reliable product. Product B is also nearly as fast as Product A ( some times even very rarely surpassing that of Product A ), and the people in the IT sector have no qualms putting this product into a personal system, or even in a production server. Product B is not without its flaws, but fewer by comparrison, and some serious flaws rarely. Product B has also pioneered several features that are still in use today by many product names, but is now out of the retail sector after one such serious flaw, and a bit of financial troubles.
So, if "review site A" gives raving reviews of a product name continually in favor despite the fact that this product will not run countless hours/days/months without crashing ( because this is not important to them ). Then of course review site A is going to get some flack for their very public words. However, ANY OS can be very, very stable with the right hardware. Yes this even includes Windows. It will probably never be perfect, but solid hardware, with solid software backing it can make a huge difference. I've been there. am there, and hoping to continue to be there in the future.
Sadly for review site A however, I take their words with a grain of salt, and have to back whatever is said on this site with countless hours of research on my end. That is ok though, I do not pay review site A for anything, and they are still very informative even if in not exactly the same way I wish.
yacoub - Friday, September 11, 2009 - link
You've certainly piqued my interest: Where can I read about Intel instability in server environments? Being a mere mortal, I only know what I read at hardware review sites like Anandtech, so I haven't heard about this before.yyrkoon - Friday, September 11, 2009 - link
Assumption is the mother of all . . . . ups.CPU's very rarely exhibit serious issues in general, and if we're going to get into that Intel has always been very solid. This also is in the context of new parts.I am sure the mention of Intel as a specific brand in this blog was a generalization. As a matter of a fact, I am a long standing "fan" of AMD. I am glad AMD is around, and in in the past have used *many* AMD CPU's ( also Cyrix if you want to get into that as well ). Now days, I actually prefer Intel CPUs. The *only* problem I have ever personally had with Intel is that in the past, their CPUs were not obtainable simply because of the outrageous cost for many people. Now days, and probably within the last 4 or more years, their prices have been very reasonable. This is not to say that Intel as a whole is not without their problems. Read: the company, not the product.
The context of my post was more along the lines of motherboards, memory, power supplies, video cards etc. And while I left out details on purpose, the example was loosely based on real events. I bet Jarred knows which products I was talking about. Because I have talked about them more than once over the last several years. Also this example was an extreme generlization.
Xortin - Friday, September 11, 2009 - link
Cyrix chips were a problem many moon's ago. Have not had a lot of problems with either Intel or AMD in a long while. Server setup or desktop.JarredWalton - Friday, September 11, 2009 - link
Actually, I'm not entirely sure what you're talking about, in part because you were so vague. My guesses:Intel IGPs still frequently suck, especially in terms of drivers. NVIDIA chipsets also have issues on a regular basis... but then, problems with drivers and chipsets have plagued pretty much every company.
As a whole, I would choose Intel chipsets over the competition, but it's a tough call. (Note: if we're talking about IGP chipsets, that recommendation would change. Sadly, good IGP chipsets frequently come with the less desirable other items, like unstable chipsets/drivers or slower CPUs.)
While I have used plenty of AMD setups in the past, I buy for what is currently the best platform. That means all of my Athlon 64/X2 systems are gone now, except for the one Athlon 3200+ 939 setup I keep around for my daughter. My main work and play systems are actually both running Kentsfield processors -- Q6600 @ 3.30GHz and QX6700 @ 3.20GHz. There's also an old Pentium D sitting in the corner gathering dust... though it has an X38 motherboard so I should probably get a different CPU from Anand or Gary.
The rest of my computers are now all laptops in various stages of testing. Two are AMD-based, and frankly they're the slowest and least desirable *laptops* by a large margin. The Dell Studio 14z is actually looking like a very good laptop that starts at $650, and $200 extra will get you pretty much everything you really need. Battery life, performance, stability, and even a 1440x900 LCD, with a package that weighs in at less than 4.5 pounds. You can get AMD-based setups for less money, but they're slower in every area and offer about 2/3 the battery life.
Totally off topic post, I know, but some of you might be interested in reading it. :-)
yyrkoon - Saturday, September 12, 2009 - link
Speaking of laptops, I purchased one recently too, and could not be happier with it. It is not a high performance jobby ( but I did look at one of those too; By Asus even ), but it also only set me back $399 , or $450 if you include the 4GB Crucial kit I purchased at the same time.The main reason I purchased this laptop is that everyone I have spoken with that has owned one of this brand has been very happy with them. But these people also try to take care of their equipment, so they may/may not be bullet proof. The one I purchased is definitely flimsy, but I do not move it around a lot, and when I do, I go out of my way to take care of it. So, it is not perfect either, but I am very happy because it does the main thing I expect from any system I own, or build. It does not crash.
Speed is important too in some cases, but when you purchase a pre built system ( laptop in this case ) with a T3400, 2 GB of RAM, 160GB HDD, and Intel 4500M IGP. What are *you* expecting ? I would also expect AMD laptops to not be top performers, but you have to know this going in. This is why lots of "research" is nearly mandatory, other wise not matter what you purchase is going to be a big let down.
So, I guess my point here is that "desirable" in this context is subjective. Myself, I purchased a laptop mainly because we're 100% solar ( off grid ), and could not afford to spend $30,000 for batteries, plus more for other equipment like panels, or wind turbines *JUST* to run my desktop. By comparison, it is a beast, and is gaining on a couple of years old now. It also sucks down 224W continuous while playing Modern games. 165W while idling. The laptop ? 40W while gaming, or 17W idle. Big difference. So I chose an Intel laptop only based on they seem to be leading the market in low power systems that are still usable for most every day tasks. Read: an atom ain't going to cut it. That, and I payed less than what they seem to be asking for a high end atom system.
yyrkoon - Saturday, September 12, 2009 - link
Strikeback hit it on the head. Asus, and ABIT. Roughly, very generalized, and blown out of proportion.I will miss ABIT sorely, but not because of the name, because of the product. I had never, ever had any product from this company that was not maintained for a long time, or at least until all the issues were ironed out. I am far from alone in this school of thought. VERY far. The products in the end were perhaps a tad slower than some of the competition ( they usually landed in the middle of the pack from memory, with a rare speed demon from time to time ), but were very solid as far as Stability went.
They were not however perfect, as no one is. I had a system board go south after 5 years, because of the substandard capacitors used. 5 years . . . and they were not the only company who fell into this pit as it were. The board was a KT7A RAID-100 board, still have it ( for parts ), but it does not function any longer despite a cap replacement.
Anyways, my point on this subject is not to glorify ABIT ( although maybe they deserve it ), or blow smoke up anyones butt. My point is that the practices used by this company should be the rule, and not the exception. Instead, we get hardware that people from reviews call stable after running successfully for a very short amount of time( and in the real world the darned things will not run a week without some form of a crash related to hardware ). This is even with stock proper settings. And it is not even the reviewers fault, except that perhaps these products get awards that they should not have, until they exemplify true ability by reputation. Enter our "budget line" product . . .
Now, this is not just a problem with motherboards. This is a problem that seems quite capable of popping up in any part of the market. Many of us know this, and avoid these companies/products like the plague. This is not because of fanboiism, this is because those of us who do run small businesses do not wish to lose business because of substandard garbage. Maybe my thinking here is antiquated . . . but you know, if it "just works" I can actually get some work/play done, instead of fixing the damed thing ( again ).
By the way, I prefer systems with Intel chipsets, and if the system is for play, I just drop in a discrete card. I am not adverse to other products, but lately the product would have to be proven to me in person. I am also pro Intel CPUs, even though I have used many AMD CPUs in the past( even before the Athlon ). Cost factor is what played the biggest role for me, because like I said in a previous post. CPUs are very rarely an issue. Sometimes different branded CPUs are slower, sure. That is a given. If the CPU exhibits serious problems, hell we'll find out about it before we need to use one ( because of sites like Anandtech ), and we can just bite on the next revision. Right ?
yacoub - Sunday, September 13, 2009 - link
Wow I didn't realize Abit was gone... just read the Wikipedia article. I was wondering why I hadn't seen a board from them in a while. Kinda sad. I used to like their boards back in the day. :( Though I did also have problems with one and that was what got me looking at other brands for my motherboards since.lyeoh - Saturday, September 12, 2009 - link
Hi, regarding the slow laptops, have you ever tried benchmarks with the NtfsDisableLastAccessUpdate registry setting set to 1?This might also affect filesystem benchmarks on SSDs.
See: http://technet.microsoft.com/en-us/library/cc75856...">http://technet.microsoft.com/en-us/library/cc75856...
The reasoning behind this is random writes on laptop drives are very slow. Random writes on some SSDs are slow too.
With the default NtfsDisableLastAccessUpdate setting, Windows will write to the disk every time it opens a file or directory.
I suspect this might affect performance a lot. Maybe you can benchmark it :). Naturally this might screw up some apps, but I have NtfsDisableLastAccessUpdate=1 for years and it's fine for my usage.
strikeback03 - Friday, September 11, 2009 - link
My guess is specific motherboard vendors, such as abit and Asus.SJD - Friday, September 11, 2009 - link
So basically, clock for clock these things run almost identical with LGA1366 processors. That's really interesting, but doesn't take away that as a stock processor the i5 750 makes a really sweet choice at a great price for those who don't want to overclock and want lower power consumptionAnn3x - Friday, September 11, 2009 - link
Thanks for taking the time to do the extra tests. Good to know youre listening.Yeah some people went over the top with the stupid irrational criticism but what did you expect with incendiary remarks like (paraphrasing) "all CPUs below the 965 are now pointless".
To me personally, that was what really annoyed me about the 1st article. A silly blanket statement ignoring many of the real world reasons for choosing a 920 or whatever.
Anyway, thanks for the additional testing. Lets hope thats the end of it regarding this debate. Anandtech has always been a site Ive admired for its high quality and insightful testing i guess to have read one review that I disagreed with in 3 years isnt really too bad!
murphyslabrat - Friday, September 11, 2009 - link
Could you give a real-world scenario involving the justified purchase of a 920? I sure as hell can't.Ann3x - Sunday, September 13, 2009 - link
Anyone who knows how to overclock?Anyone with SLI?
Anyone with CF?
Anyone with a PCI-E Raid array
Can you think of any good reason (apart from ineptitude, laziness etc) why these people would choose anything BUT a 920?
TGressus - Monday, September 14, 2009 - link
It really boils down to the target markets. In the consumer space, the 1366 platform is a luxury and 1156 has proven worth the wait. That being said it was marketing genius to put out the top gun platform first and bilk the early adopters.I commend the earlier poster for identifying his present needs, and contrasting that with the relatively short tick-tock production cycle Intel pursues. Any kind of long term investment in a desktop Intel chipset is a financial mistake now, and a technological mistake in 18 months.
Other market space could be interpreted differently...
has407 - Friday, September 11, 2009 - link
Because it's the cheapest entry point for LGA1366/X58, which allows for a wider upgrade path than LGA1156/P55. E.g., memory (24GB vs. 16GB); ECC (possible vs. never); cores (6-8 vs. 4); PCIe lanes (36 in various flavors vs. 24 in few flavors).Is that potential upgrade path worth the premium? Depends. If you're looking at longer term system-level price/performance, then arguably yes. I was waiting to see what the new parts offered before making a decision on building a new home server, and for me LGA1366/X58--at the cheapest I can get into it, which means a Core i7 920--is still attractive.
erple2 - Saturday, September 12, 2009 - link
Curiously, I thought that too. However, the more I thought about it, the more I realized that they're not really points at all. I've had this Core2Duo since 2007, and it's performed exceptionally well for what I've wanted it to do. Now that I feel that I may want to "upgrade", my options are Core2Quad, 1156 or 1366. Well, The Core2Quad is more or less "useless" right now, given how well even the 1156 runs. So having the chipset (in my case, a P35) upgraded means nothing to me.So now I look at upgrade time (it's been 2 years, so I figure it's about that time again). So my choices are the 1156 or the 1366. Do I really think that I'm going to need more than 8 gigs of RAM right now? No, I doubt it. And that's on the cheap side. I figure that when I'll need more than 8 gigs of RAM (12, 16 or 24 gigs), a 4 Gig stick of DDR3 will be "cheap enough" to be palatable. So the real question is, will I need more than 16 gigs of RAM? I honestly can't imagine needing to more than quadruple my current memory usage. I don't even use the 4 gigs that I have now. I think that I'd be happy with 2 gigs, now that I think about it.
So, "future proof" I think is a bit of a misnomer. How long can we reasonably expect the 1366 to be relevant in the Market place? My old P4 was one of the socket 478 chips. The "upgrade" for that ended with the P4, 3.06 GHz (with hyperthreading). Needless to say, it wasn't worthwhile for me to actually spend money on the CPU upgrade. I'd question the logic of making a "future proof" decision, at least based on CPU capabilities.
The possibility of 6 cores or 8 cores might be interesting. It might be interesting if we had a significant share of applications that took advantage of more than 4 cores (or 8 "virtual" ones). I don't have any that are appreciably good at that. Most of those (video encoding) will still be stunningly faster when run over a GPGPU calculation (you can't beat 800 stream processors, or whatever NVidia has right now).
That leaves just the PCIe 2 * 8x or 2 * 16x issue. I don't run dual graphics cards, so that option isn't very useful to me either. The idea sounded good that I could, in the future, buy a second card when my single card "got slow". However, I'd just as soon buy a nice, new card that performed on par with the 2x cards, so that buys me nothing. Particularly given that finding the same card would be harder to find since either the company has re-spun a new rev of the cards that aren't 100% compatible in SLI or Crossfire, or that they're just not made any more. So the cost of the card wouldn't be that much cheaper.
I dunno. I always thought that "future proofing" a purchase sounded good. However, I've honestly never actually taken advantage of that. When things get too slow (and more RAM thrown at it doesn't solve the problem), I usually just buy the new big 2 (CPU which usually begets a new chipset, hence also buying a new MoBo). Every now and then, I can use the RAM from one base computer to the next.
95thRifles - Friday, September 11, 2009 - link
Appreciate the effort,any chance of maybe some old scores from a C2Q maybe even C2D etc for some comparisons, to see if it would be worth people updating their existing systems?ClagMaster - Friday, September 11, 2009 - link
I second that motion. I too appreciate the efforts.I would like to see some PCMark and 3DMark scores for an E6600 and a Q6600 included in these reviews so I can see how they compare to i5 and i7 LGA 1156 processors. Many of us mainstream people will be upgraded from these processors and want to know if its worth it. If that does not happen, then the only other alternative is to visit Toms Hardware CPU charts and scale with the i7 920.
Gary Key - Friday, September 11, 2009 - link
Sometime between midnight and four am I plan on running a C2Q system for comparison. :) That delay is my fault. We had only planned on showing the P55 and X58 board results in the initial motherboard articles. We added the 790FX/965BE earlier this week and now if everything works out properly, we will have a P45/C2Q setup in the motherboard articles based on a large number of requests. However, for a quick CPU comparison, check out the Lynnfield launch article.yacoub - Friday, September 11, 2009 - link
Gary, you are a huge part of what makes this site the valuable resource it is. Thanks for going the extra mile to make it comprehensive.justme2009 - Friday, September 11, 2009 - link
"I will interject a personal note here, the emails/private messages that outlined a strong case for additional research and testing certainly held a lot more weight than comments like "You are Intel's payroll...", "Worst review ever...", and the moonshot , "Illegal benchmarking methods..". First off, if we were on Intel's payroll we would not be working here (a logical conclusion right? ;) ) As for the other comments, everyone is entitled to their opinions. We do our best to keep an open forum and let the comments fall where they may, but offering constructive criticism and facts to back up those comments is what actually causes change, not endless shock posts or attention grabbing statements. I still have hope in people abiding by the rules of Internet Etiquette, but apparently we are still a long ways off from that happening. I will step off the soap box, well, until the next article...."Gary, being a popular website on the net, it goes with the territory. Atleast with Anandtech, you are open and entertain other people's suggestions and opinions. I think that sets you apart from the rest. Keep up the good work!
The comment I made about this latest processor is, it's not a big leap by any means, it's really nothing to write home about and I felt AT was overhyping it a bit. I think the bigger leap (as I said in the other article) will be mobile Nehalem and beyond, and that will be an article worth reading.
happyrock - Monday, September 14, 2009 - link
I'm not sure if this will work here...but as a staff member at geeks 2 go we can set in our profile blocked posters ...people that we don't even see their posts...EVER...it would help keep the replies on topic or at a minimum not waste any time having to wade thru the BS posts by anyone we consider not worth reading anything they have to say...just my 2 cents worth...jasonbird - Friday, December 25, 2009 - link
http://www.socheapwholesale.com">http://www.socheapwholesale.comhttp://www.1stjerseys.com">http://www.1stjerseys.com
http://www.socheapwholesale.com/NFL-Jerseys.html">http://www.socheapwholesale.com/NFL-Jerseys.html
http://www.1stjerseys.com/NHL-Jerseys.html">http://www.1stjerseys.com/NHL-Jerseys.html
Eeqmcsq - Friday, September 11, 2009 - link
You guys should just ban snakeoil. All he does is make outlandish and idiotic claims just to get a rise out of other commenters. He does this too at Techreport. Now he's going to think that he's justified in all the noise he tries to stir up. It really ruins the more intellectual atmosphere I've come to expect out of Anandtech comments.Anyway. Personally, the test combo that I'm more interested in is testing with Turbo off as well as on. Since the performance boost provided by Turbo is variable based on whatever other apps the user or operating system may be multitasking in the background, testing with Turbo off provides the "guaranteed minimum", i.e. a baseline level of performance that a Turbo capable CPU will provide, and that any Turbo boosted extra is considered a free bonus.
Otherwise, people who skim through the graphs may think that the CPU will ALWAYS provide the Turbo boosted score, instead of understanding that the Turbo boosted score is an "ideal" score, when conditions of the CPU are at its best for Turbo.
Anyway guys, keep up the good and hard work.
Inkie - Sunday, September 13, 2009 - link
"You guys should just ban snakeoil. All he does is make outlandish and idiotic claims just to get a rise out of other commenters"Then they'd have to ban TA152H and others for the same reason. Just let them speak. I think it is an error for Anand or Gary to reply to them though. That just gives them status in their own eyes and encourages them to continue.
TA152H - Sunday, September 13, 2009 - link
Why are you bringing me into this?Actually, I back up my clains, unlike what you did. I don't just say this or that sucks, or was done wrong. I tell them why, and back it up.
Gary can complain about shock posts, but, I put more effort into my posts than they put into their P55 motherboard pictorals :-P. And I DO back them up.
Most things in life are not black or white, good or bad, but a shade of gray. People are simpletons, and like it simple, but it's just not how most things are. P55 isn't a perfect product, by any means, and it's not all bad. If you don't have people pointing out the flaws in the methodology, then you have holes in your information. Feedback is not only acceptable, it's necessary.
More to the point, when you idiots were all pointing out how good the performance was on the first P55 article, I was explaining the memory performance was abysmal for what the product was. No one else seemed to see it. Guess what? I was right. The numbers were improved for the released product (and I kept saying I thought it was from pre-release hardware), so the terrible performance didn't make sense after all. Did the writer see it? Nope. He was saying how it was the greatest thing since cheddar cheese? Did the masses see it? Nope. Like proper cattle that can't think for themselves, they just ate it up and did as they were told, so to speak.
You need critical people that are going to look at what you do, and take you to account when you do something wrong, or just don't do something. I need it when I do things to. I don't always like being shown it, but, at the end of the day, the end result is better.
That's really what they should be here for, right?
Inkie - Sunday, September 13, 2009 - link
Actually I've hardly made any posts on Anandtech in the past. I've just read your particular brand of posting recently, since I was interested in the 1156 platform. You have said many incorrect things about the 1156 platform, both before and after release, using ridiculous, hyperbolic language to do so and what's worse, you never acknowledge your mistakes and habitually attack others. You are no better than someone like snakeoil. Actually, snakeoil may be wrong, but he is a bit less free in insulting others. I think you are funny sometimes though, just because you can be so wrong and technically incompetent. Writing this post, I still have to chuckle eg. about what you said about pin-multiplexing on the DDR3 and PCIe interfaces. I don't want to start a flamewar, so I won't reply any further. No point anyway.TA152H - Monday, September 14, 2009 - link
I read stuff from little losers like you, and I'm wondering if you even have the common sense to realize the hypocrisy in what you're saying. Are you really that stupid? I guess you are.You insult me, and then whine that I insult people? Duh. What a moron.
You say I was wrong about things, but I don't see a single quote.
You have heard about multiplexing pins right? Processors have done it before, in fact the 8086 did it with the address and data bus. Intel shaved off 210 pins. I haven't seen the pinout yet, but it's probably out there. Oh, and I never said they multiplexed the pins, I said it's a possibility they did, and it would be nice if we got information like that instead of pictures of motherboards.
Keep it straight, OK loser?
Inkie - Monday, September 14, 2009 - link
There is no multiplexing of pins across the PCIe and DDR3 interfaces, I can tell you. I'm not going to waste time quoting you. Notice the only thing that could really be interpreted as an insult is that you are sometimes funny by virtue of being so wrong headed. Your mistakes, hyperbolic language, inabality to admit your mistakes, attacks on others or whatever are just facts that anyone can see.So, I did reply further. But, what do you care, I'm a loser and a moron, right?
Inkie - Monday, September 14, 2009 - link
Just a further expand the point: you realise that the DDR3 and PCIe interfaces are separate interfaces, routed to different places on the motherboard? I was really laughing that you would even make such a speculation.Inkie - Sunday, September 13, 2009 - link
"Anyway. Personally, the test combo that I'm more interested in is testing with Turbo off as well as on. Since the performance boost provided by Turbo is variable based on whatever other apps the user or operating system may be multitasking in the background, testing with Turbo off provides the "guaranteed minimum", i.e. a baseline level of performance that a Turbo capable CPU will provide, and that any Turbo boosted extra is considered a free bonus."Turbo off doesn't really give a "guaranteed minimum". If there is heavy mulitasking then performance on a given task can be much worse than the turbo-off score for that task measured in comparitive isolation.
Eeqmcsq - Sunday, September 13, 2009 - link
Yes, it's true that extremely heavy multitasking can overload a quad core CPU with too many threads, thus degrading performance for all tasks, but that's not the fault of the CPU. That's simply more tasks than the CPU can handle. Also there's the side issue of disk contention as multiple threads fight for access from the hard drive. That could also affect a benchmark, but that's also not the fault of the CPU.But I'm thinking of cases where the background multitasking pushes 1 or 2, maybe even 3 cores up to a constant and prolonged 100%, but the quad core still has enough "room" to theoretically run a single threaded benchmark without being affected by the background tasks.
For example, suppose an instance of Firefox and an instance of Openoffice get into a buggy state where each are running an infinite loop, thus is not causing disk contention. In this case, 2 of the 4 cores will reach a prolonged 100%, but there are still 2 cores available for a single threaded benchmark, such as Cinebench. On a Core 2 or Phenom II, there should NOT be a drop in Cinebench performance. But on a Lynnfield, Turbo either can't kick in, or will kick in with much less of a boost, so there SHOULD be a drop in Cinebench performance. Assuming all other conditions aren't otherwise affecting Cinebench, the performance drop should theoretically be no worse than a Turbo-off test. That's what I'm referring to about a "guaranteed minimum".
This is also why I suggested a multi-instance tests of single threaded benchmarks. That way, we can see how much of a drop off there is for 1x, 2x, 3x, 4x concurrent instances of single-threaded-Cinebench both in non-Turbo CPUs of Core 2 and Phenom 2, and also in the Core i5/i7/etc. Hopefully, Gary has read my suggestion and added such a test.
Inkie - Sunday, September 13, 2009 - link
So, what you are essentially saying is that the cores have to be loaded just right: not too much and not too little, for the non-turbo score to be really representative of actual performance in a given task. For me, that's trying too hard to make it interesting. Not that I'm criticising your interest. I just don't really share it.Eeqmcsq - Sunday, September 13, 2009 - link
That's sort of what I'm saying. I'm interested in getting a scoring "range" that Turbo capable CPUs can provide to give a more complete picture. It's kind of analogous to how hard drives have a read speed range, example: 60MB to 80MB, depending on whether the data sits on the inside or outside areas of the platter. In normal usage, you're going to get various read speeds, but you won't read any worse than 60 MB/s, assuming no other external factors, but you also won't ALWAYS get 80 MB/s either.Likewise, Turbo capable CPUs have a performance range from no Turbo to best possible Turbo. That's why I'm interested in the no Turbo scores, to get the bottom end of this range.
Inkie - Sunday, September 13, 2009 - link
Look, you can get a range of performance from crawling along because of other things go on to performing at maximum with turbo-on (and reaching the maximum level of turbo) relatively isolated. I use my computer in the real world and don't choose combinations of tasks to make the turbo-off-relatively-isolated-figure especially significant. It might happen by chance that I get something like that sometimes, but it is no more interesting or significant than anything else.Just about the only useful thing that sites can benchmark is the maximum performance that can be expected. It is understood that if you multi-task you can have a whole range of lesser performances for a particular task, depending on how heavy you multitask. The turbo-off-relatively-isolated figure is just an arbitrary figure plucked out of that range that might happen by chance sometime. It's pretty much a waste of time benchmarking it if the goal is to give people an expectation of what they will get in the real world.
Mr Perfect - Saturday, September 12, 2009 - link
That's a reasonable request, I'd be interested in seeing that too. Like Eeqmcsq says, Turbo is like a free bonus.Also really looking forward to the expanded i7/860 tests. In the original article, it looked like a great balance between the pricey models and the cheap alternatives. It's a shame you didn't have more time with it then.
strikeback03 - Friday, September 11, 2009 - link
When looking at their review of the ASRock X58 Extreme I saw that in the comments snakeoil (the site's biggest AMD fanboi) and SiliconDoc (the site's biggest Nvidia fanboi) went at each other.Gary Key - Friday, September 11, 2009 - link
I am running the Turbo off numbers on the 750 today, which means getting these motherboard articles out is taking just that much longer. ;) However, we understand the importance of the request and I will show a few results late today. Anand will followup next week with a full analysis. Thank you for the suggestions.Eeqmcsq - Friday, September 11, 2009 - link
You, sir, gets a THUMBS UP! Thank you.Other suggestions I thought of that you might try for Turbo capable CPUs: Run a varied number of concurrent instances of a single threaded test. For example, run 1 instance of a single threaded test, then run 2 separate instances of the test at the same time, then 3, then 4. See what the numbers look like and how much Turbo contributes based on the # of busy cores.
yacoub - Friday, September 11, 2009 - link
You know, that's a really good point about a "guaranteed minimum". I hadn't thought of that when i defended the use of testing with Turbo on "because that's how it comes out of the box". I can appreciate your reasoning and agree with you that it would be helpful to also have Turbo Off numbers for the benchmarks. Looks like Gary will be doing that, so that's cool.Also, Gary is awesome. :)
jonup - Friday, September 11, 2009 - link
Gary, can you also add some results of OC with Turbo On. I know that the final OC will be lower, but in the initial review Anand pointed out that in a lesser threaded circumstances the effective maximum overclock with Turbo On is higher (i5/750 and i7/860) than the maximum OC with Turbo Off.The benfits of running lower OC with Turbo On will not only be in single and dual threaded app's but also power consumtion => lower temps. Good for 24/7 OC.
I am intrested in the socket 1156 and I run my systems OC 24/7.
Thanks Gary!
Gary Key - Friday, September 11, 2009 - link
I have been working on that also. :) I just received another retail 860 today that on a clock for clock basis needs about 0.06V less than our first sample, so the results should be more in alignment with the retail processors on sale now.Nich0 - Friday, September 11, 2009 - link
Regarding Anand's OC numbers (with Turbo) in the main article, something is off for the 860 and the 750. I posted in the comments to the main article (around page 32/33!?!) about it. For example the Turbo 1C number would be 26 BCLK, ie a turbo of 6. Am I missing something?jonup - Friday, September 11, 2009 - link
Thanks buddy!goinginstyle - Friday, September 11, 2009 - link
If the rest of the results are like this one then I have a bad feeling the comments are going to get worse. These results do not bid well for those who asked for this information. Pulls up a chair and grabs the popcorn, just waiting on SnakeOil and the other nut jobs to show up.snakeoil - Friday, September 11, 2009 - link
here we can see that a phenom 2 965 annihilates the lynnfield 750 at stock speed, that is lynnfield without overclocking.and this is just one test, remember that this render test favors intel. there are many others tests where phenom 2 beats lynnfield by higher margins.
now the truth is surfacing.
we the people demand that you repeat all the tests of your lynnfield review without the auto overclocking at stock speed.
just to be fair..
Fohtey - Tuesday, September 15, 2009 - link
I'm not sure 19 seconds is an annihilation, but at "stock" speeds the Phenom II 965 does do better than Lynnfield 750 at stock.1) Aren't "stock speeds" determined, in part, by where the manufacturer sees there product's niche? In the past, parts of processors have been disabled for the very same reason. So it stands to reason this may be the case here. I guess we'll see. I recall a feature or two offered in the i7 series are not available in the i5. Again, I'm guessing it's marketing IMHO.
2) These tests determine a processor's capability. If you're complaining that we don't see "apples to apples" comparos between AMD and Intel, it's because Intel processors are more stable when over-clocked at higher speeds. So if AMD is competing with Intel for fastest processor, AMD is likely to set its "stock speed" closer to its stability threshold than Intel will with theirs.
So insisting that "stock speeds" are a relevant measure is like comparing two runners when they're jogging and determining who would win a given race. It's only when they're really pushing the limits of their abilities do we see who the best runner really is.
It is the same here. So based on the information we see, i5 clearly is a better performing processor for this application.
These arguments are only relevant to us system tweakers anyway. The average Joe couldn't care less. All he's interested in is realized speed and not the barely perceptible differences between two closely matched processors.
poppyharlow - Monday, September 14, 2009 - link
Hi,Could you explain to me how you call it fair when you compare
the Phenom II 965 BE which runs at 3.4Ghz (stock speed) to the Lynfield I5 720 that runs at 2.66Ghz (stock speed, turbo off) which
makes it 27% (740Mhz) slower in terme of clock speed?
If you take this clock speed into account (just to play fair like you want) then we should compare the PII 965 BE to a Lynfield I5 720 overclocked to run at 3.4Ghz also right?.
If you apply this 27% difference in clock speed to the Lynfield I5 750 (with turbo mode off) then I'm not quite sure that it's still slower than the PII 965 BE...(245 - (245 * 27%) = 178 compared to 188 on PII 965 BE).
Please explain.
Regards,
maxxcool - Monday, September 14, 2009 - link
Do you realize what you just sail OilySnake, you just said:"our biggest most expensive chip can beat up you SMALLEST SLOWEST CHEAPEST CHIP"
Thats also like saying :
"My 12grader can beat up you're toddler"
Do you LIKE beating toddlers? is that what you are saying OILYsnake?
Your just as much of a troll and waste of skin here as you are on Techreport. Why don't you step into traffic and do the world a favor?
The tests are 100 VALID.
At stock settings, the tests reflect exactly what every intel user will get. if you dont like that.... stop reading the reviews you pathetic little meat sack.
or better yet, go design a CPU... that is if you can leave your mom's basement long enough....
OH and don't forget you IDIOT, you are the moron that was swearing up and down that I5 would not support virtulization...
tool.... no wait... your not good enough to be a tool.... your a tool-bag.
Anand, Gary.... ignore this idiot. talk to damage on Techreport... this guy is a complete douche.
SlyNine - Friday, September 11, 2009 - link
This is stupid, The 750 with turbo boost is NOT an overclock. Its simply how the CPU manages thermal distribution. You may not be happy with the results. But until you can provide you're own data that is repeatable and doesn't cripple Core I5/7 CPU's, by say disabling turbo boost. Then you're just grasping at straws.Face it AMD cannot currently compete in the high end. Get over it.
You make some pretty far fetch accusations, with no backing. At this point you're just trolling. It's to bad I fill the need to feed your ridiculous comments.
snakeoil - Friday, September 11, 2009 - link
turbo is overclocking. if you are going to benchmark using an overclocked processor then that is cheating. and that's illegaldo you think people is stupid?
Kaleid - Sunday, September 13, 2009 - link
Sigh. There is nothing illegal about it. Its a built in feature.If the turbo mode is a success expect AMD to do something similar.
ssj4Gogeta - Saturday, September 12, 2009 - link
It's what performance you're getting for the money moron, not what the processor is doing to get you that performance.Mattus27 - Friday, September 11, 2009 - link
Turbo is overclocking?It's an Intel-endorsed feature that's enabled by default, you ignorant moron. If you had brains, you'd be dangerous.
Etern205 - Friday, September 11, 2009 - link
Let me just say that I've build quite a few of AMD's based system going back to the K6-2 era and they are pretty good processors, but your comment is just simply idiotic.And I bet if AMD has this feature, you'll say it's a fair comparison.
GET A LIFE!
fantoosh420 - Sunday, September 13, 2009 - link
be it OVERCLOCKING (according to you) or an innovation called turbo boost (according to rest of us))oops... this line was meant for snakeoil.
fantoosh420 - Sunday, September 13, 2009 - link
true... i agree...we don't hate AMD. they are pretty good CPU manufacturer.... but having said that, we are also end user... and all of us want the worth of our every single penny which we spend on CPUs. now there are ways.... poor people who can't afford a $280 cpu + $150 motherboard + $120 RAM and so on, will go for a AMD. it gives nice performance for its price. People who can afford that much cost will definitely go for intel.
At the end of the day, it is your pocket which becomes deciding factor rather than what really you want. I want to buy a Dual Xeon 5580 based system but can i afford it ?? NO. so what do i do ?? may be go for a Xeon 3550 or something... can i afford that too ?? no. so what do i do? may be go for a AMD 955BE or something....doesn't matter what i want. but if i can't afford it i'll choose the next one downside the list until i reach the affordable limit.
in current world situation, most people can afford a good $1000 system, since in that range, i7/860 gives best performance (no matter how that performance is achieved... be it OVERCLOCKING (according to you) or an innovation called turbo boost (according to rest of us)). the bottom line is what Gflops people are getting for the money that they are going to spend and as long as intel is fitting in that meter better people will go for it. when the time will come (and it has come in past once), people will go for AMD as well.
jonup - Friday, September 11, 2009 - link
...people are stupid, maybe?Btw cheating is not illigal. If it was, you could take AT to court and win (if they were cheating).
After all, do I need to remind you that the 965BE is an OC 955BE?
And do you really want to see OC (24/7 or max) 965BE destroyed by and OC (24/7 or max) i5 750.
JackPack - Friday, September 11, 2009 - link
If you think Turbo Mode is overclocking, then you might as well point to the sun and say it's dark.wizzlewiz - Friday, September 11, 2009 - link
do you think people is stupid?I believe this statement speaks for itself. However, turbo is an innovation designed by Intel to increase performance. If that is cheating, how have we made it as far in the computer industry as we have? It is like comparing a single core to a dual core, and saying for the test to be fair that the dual core has to shut off one of its cores.
densetsu612 - Friday, September 11, 2009 - link
... and what about the auto under-clocking and under-volting when the Lynnfield CPU is running idle?perhaps the Phenom II BE needs to be undervolted and underclocked when it is running at idle to measure power consumption.
(just to be fair)
jonup - Friday, September 11, 2009 - link
Wow, you are so predictable.BTW, are you referring to the "turbo on" as auto OC? In that case you want to compare a Stock 965 to cripled (worse than stock) i5. I mean, do you prefer an article that says "Intel today introduced its new attempt to detrone AMD from its dominant position. However, the new i5 is a more expensive, slower processor that is waste of time testing at stock configuration (or overclocked to stable 24/7). Instead of wasting your time reading a pointless article, just go get yourself 965BE and live happily ever after."
If you ask Gary, I am sure he could e-mail you this massage every morning and you would not have to ever come to AT or any other site on Intel PR.
Just before you go all out with your AMD Fanboyism at me, all three of my computers have AMD CPUs. (I built two and the third is a laptop).
snakeoil - Friday, September 11, 2009 - link
only by doing this you will prove that you are not on intel's payroll.snakeoil - Saturday, September 12, 2009 - link
we also want to see temperatures when the processors are overclocked because some reviews say that when overcloded lynnfield reaches 95 degrees centigrades even when they were using a Thermalright MUX 120 which is ridiculous.http://www.guru3d.com/article/asus-p7p55d-deluxe-m...">http://www.guru3d.com/article/asus-p7p55d-deluxe-m...
DigitalFreak - Sunday, September 13, 2009 - link
We? Are the voices in your head speaking to you again?Eeqmcsq - Sunday, September 13, 2009 - link
It's the voices that say "We are Borg".maxxcool - Monday, September 14, 2009 - link
No its the voices that say "Jerk off the the Guild music video with Felicia Day in it"....I have deveopled quite the hatred for this open-palmed douce tool-bag...
snakeoil - Saturday, September 12, 2009 - link
we also want to see the power consumption when the processors are overclocked because some reviews say that lynnfield power consumption skyrockets when overclocked.http://www.guru3d.com/article/core-i5-750-core-i7-...">http://www.guru3d.com/article/core-i5-7...re-i7-86...
maxxcool - Monday, September 14, 2009 - link
oh yeah, and AMD is sooooo special you moron. did you not read Garys post you complete tool-bag douche drinking tampon stir stick?the 965 will have a over 200wat tdp at 4ghz. you cant run that in a common consumer system.
You want fair... heres fair. overclock both on LN2. see who wins. done. oh wait its been done and your sphincter is in ruins....
Go back to your moms basement you idiot.
Lifted - Saturday, September 12, 2009 - link
I guess they should disable SpeedStep/PowerNow! in the mobile processor reviews too, you know, to test battery life at stock speeds without "cheating".Alastayr - Friday, September 11, 2009 - link
Oh snakeoil, I love you like a father must love his constantly underachieving, yet futilely continuing second born son. Always trying, never giving up and sure as bloody hell fighting till the last breath.You're the only reason I read AT at all.