i5 / P55 Lab Update - Now with more numbers

by Gary Key on 9/15/2009 12:05 AM EST
Comments Locked

77 Comments

Back to Article

  • pnot00 - Friday, September 25, 2009 - link

    Where can i get the 2x4GB sticks in this article? I can't find them anywhere and would like to put 16GB in this board for a small server.
  • pnot00 - Friday, September 25, 2009 - link

    duh n/m i must have been hoped up on too much coffee... 4x2GB.

    I really need to find 4 sticks of 4GB and the only ones i've found are by Kingston and they want $1,000 for them.

    Any suggestions?
  • FazliAmri - Wednesday, September 16, 2009 - link

    Hello.

    I'm eagerly awaiting for the finished review. i'm currently upgrading my system, thinking of 920/x58 or 860/p55, been scouring the net for the information that'll affect my decision. i think i'm on the verge of choosing but the thing that 'i think' is missing is

    cpu temp idle/load at OC 3.8;
    power consumption idle/load at OC 3.8;

    if you could include this data i thank you.
  • TA152H - Wednesday, September 16, 2009 - link

    You'd have to be retarded to buy a Lynnfield if you have enough money to get an i7 920 setup.

    This site tries very hard to push the P55, while trying to appear unbiased. Why? Who the Hell knows? Maybe Anand decreed the P55 should be put in a good light. I really don't know, but at this point I'm sure they are being biased, either consciously or unconsciously.

    But, you're really better off with Bloomfield if you can afford it.

    In this article, they make the overclocked processors look similar, but what they don't tell you is they run the memory controller faster on the Lynnfield, as well as the L3 cache. Why don't they tell you? They don't want you to know.

    So, clearly, the Bloomfield is faster even on the benchmarks they selected. I suggest you go to better sites and see the actual benchmarks, where you can see a slight performance increase on virtually any workload when the comparisons are apple to apple.

    You get better expandability, better granularity with uncore clock speeds, and memory (since you can add in multiples of two or three), better performance, and a better upgrade path since you will be able to use multi-GPU cards without penalties, and will be able to upgrade to a 6 core CPU without yanking the motherboard.

    Just make sure you get the D0 stepping of the 920. The C0 is slightly slower, and uses more power.

    If you can't swing the price of the 920, probably the i5 750 is a decent processor. I wouldn't even think about the 860 or 870. They run into Bloomfield territory, and why suffer a brain-damaged platform when you don't need to?
  • erple2 - Thursday, September 17, 2009 - link

    Curious. I don't see any actual difference between the 920 and the 860. I wouldn't consider anything under 3% to be significant, however.

    You're speculation on the bias of one site or another is just more bias, but on your part. I don't (and I'm pretty sure that you don't) know nor can prove (and neither can you) prove any bias one way or the other. From the benchmarks that I've seen from this site, it's clear to me that the P55 is a strong platform and competes well against the 920.

    Who cares if the memory clock runs faster or slower? How much of a significant difference does it really make to the end user? I suspect that the answer is "No way to be noticeable".

    I doubt that if you're concerned over saving a couple of bucks on the CPU/Mobo, you're not going to see any upgrade path along the X58 as an actual viable alternative. I seriously doubt that the 6 core processors will ever get to a "reasonable" price that will make it worthwhile to upgrade to, particularly if you're thinking about going the 920 route rather than a much more expensive CPU.

    My prediction (and it's just that, a prediction) is that when the 6 core processors are "mainstream" (which Intel hasn't really suggested that they will be in the near future), the "next best thing" (i9? whatever the future Intel line is) will be out, and prices on those will make the 6 core upgrade path not worthwhile. All while using up more power without getting any tangible benefit in the short or long term.

    You can argue that the memory controller is "brain-damaged", but from what I've seen from benchmarks here, there doesn't appear to be any tangible problem with that. Oh, a tangible difference would be in a real world scenario, not any synthetic memory stress test benchmark.
  • TA152H - Thursday, September 17, 2009 - link

    You're not really comprehending what I'm saying.

    Of course this site won't show any appreciable difference, that's my point! They skew everything so it won't.

    For example, did you know they had to set the 860 to 1.312 volts, while they set the i7 920 to 1.232. Hmmmm, you probably didn't.

    Keep in mind, if you increase the uncore speed for the processor, you not only improve memory performance, but you increase L3 cache performance, which is important.

    I've already given reasons for my comments, so don't say it's just speculation. You think running the uncore slower, without mentioning it, is by accident?

    Look at other sites that really testing, and you see on a lot of benchmarks, the performance spread is different. This site will hide pertinent information, and then show benchmark results, knowing that most of the people won't really look at the details. And then they can say crap like, the difference wasn't really important.

    You really don't think it was worth mentioning they were running the uncore 200 MHz lower? Or the processor required considerably less voltage to reach the same clock speed? This is computers we are talking about, and computers are precise. You can't come to conclusions without precision in how you test, but, strangely, they have become really sloppy, and all in a way that benefits the P55.

    It's really bizarre. Not that I mind it, I'm past being irritated, and now I'm fascinated by just how gullible people are, and how unquestioning. People don't question what they want to see. It's been very educational. The sad part is, the ones who object to results are mostly doing it on partisan grounds (meaning, they are pro-AMD), although it's always refreshing to see others who realize the i7 920 is the best in it's price class, despite the disinformation spread by this site.

    I don't think Intel made a mistake with the brain-damaged Lynnfield, by the way. I think stock, which is how most processors are sold, it's fine. Well, I do think they made a mistake; not adding a IGP with it. I realize it's more difficult with an IMC, but they've had plenty of time to get something out. So, it's really a nothing technology that will fade into obscurity until they come out with an IGP, or get rid of the i7 920. It's brain-damaged, so it's not going to attract the high-end buyer, but it also doesn't have a proper platform to be a true Celeron. So, I think it's going to sell poorly. I don't think they'll be very successful conveying the value of it, although I think the i5 750 is best in class for that market segment.

    I think AMD's new Athlon will sell really well. It's quad core, and cheap, and runs at reasonable clock speeds. If you're the typical buying public, you see quad core, a lot cheaper, with higher clock speeds than the i5 750. The nice thing is, it's actually smaller too, or reusing dead parts that would have to be thrown out. For people that have a clue and want high end, it's the Bloomfield. I just don't see this having a big market near term.

    Intel is a screwy company. They get some things so right, like the Atom, and screw up things like not getting it a good chipset. They come out with the Lynnfields, and don't give it a proper IGP. Most computers DO sell with IGPs. Why leave themselves out of this market with these new products? I just don't understand them.
  • erple2 - Thursday, September 17, 2009 - link

    No, I do comprehend what you're saying. I believe that what you're saying isn't really all that relevant in the grand scheme of things. I'm just not seeing any tangible results (>3% difference) in actual usage.

    And fundamentally, that's my point. I've been following the computing industry since before Anand had anandtech, and one of the things that I have noticed is that synthetic benchmarks are, for most purposes, worthless. Who cares if CPU A has 30% more memory bandwidth, if the vast majority of tasks show an insignificant increase in performance in a game, or in encoding, or in compiling, general computing, or in some combination of those four?

    The point is that you're complaining about, what I consider to be, more or less trivialities. I say this because based on the numbers of a non-overclocked system, Lynnfield is exceptionally competetive with the 920. Sufficient in my mind to make the 920's really not that relevant. In fact, if you don't overclock, the 920's are actually slower in the setups that I care about (ie more or less everything except some memory benchmarks that don't ultimately mean anything). All that I can say about synthetic benchmarks is that some thing is faster than another. In fact, in those synthetic benchmarks, you can't really make meaningful conclusions as to the degree to which one product is faster than the other in a real-world scenario.

    I used to be like you. I used to put too much stake in a synthetic benchmarks. I used to think that they were important beyond making some generalized comments as to the which product was faster than another product. However, after about 20 years, I finally started to see the light.

    However, I'd like to see where you believe this to be an important distinction. I've looked at a few other sites, and haven't found any tangible differences between the two. I've looked at this site, tomshardware, xbitlabs, and pcper. Where else are you seeing such a marked (and meaningful) difference in performance?
  • FazliAmri - Thursday, September 17, 2009 - link

    hello.

    We all are trying to find " tangible differences " between 860 920 p55 x58 750 i5 i7 etc etc..

    i don't think wanting to know the CPU Temperature and CPU Power Consumption at Overclock Speed of 3.8 Ghz is "More or less trivialities". because this is My "real-world scenario"

    i'm deciding which to buy, 860 or 920. these are my deciding factors



  • the zorro - Thursday, September 17, 2009 - link

    when overclocked to 4 ghz lynnfield temperatures are almost 100C, which is catastrophic.
    and the power consumption skyrockets.

    people is already complaining.

    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...
  • strikeback03 - Thursday, September 17, 2009 - link

    For all your complaints about them hiding information and such, do you actually have any references to prove that there is more than a few % difference in multiple real-world applications between the 860 and 920 when everything is set as equally as you want? From the memory scaling articles I have seen you need large changes in bandwidth or latency to get noticeable improvements in real applications.

    The difference in voltage needed to OC is of much more interest to me. Gary mentions that the 860 might save a lot of power, which seems counterintuitive to needing more voltage. The 750 is meaningless to me, I can afford either the 860 920 platforms, and have no use for better SLI/CF performance, so if performance is equal than other factors like power/temps and stability will make my decision. Al else equal, I'll take the 920 on the extremely slim chance a Gulftown will actually be cheap enough to buy. If the 860 really does draw only ~2/3 the power, I'd go with that.
  • glenster - Wednesday, September 16, 2009 - link

    Another consideration for comparison is that the Core i7 960 3.2GHz is due Q4 2009 for $562--see Wikipedia > Nehalem > 45 nm processor architecture. Another site says it's due Oct.18.
  • grimpr - Tuesday, September 15, 2009 - link

    Its always fun to watch cases of reversal...in this case, that the AMD/Nvidia combination is proven superior to the gamers and the oems market favorite Intel/Nvidia combo. We can sense the suspicions, fear and doubt on Garys comments and we wait more from him for our further enjoyment. Gary, keep on the Intel crusade...
  • Gary Key - Tuesday, September 15, 2009 - link

    There is no fear or doubt, just a quest to figure out why there is such a difference between the NV and AMD GPUs on each platform. If you are an Intel crusader I guess your suspicions would lead you to believe that NV is dogging performance on the Intel platforms.

    If you are an AMD crusader, you would think AMD is dogging performance on their own platform. Think about that for a moment. I seriously doubt either company is dogging it as they both depend heavily on GPU sales for their survival.

    Please read this and just about any other article I have written on AMD and tell me if you really believe I am on an Intel crusade - http://www.anandtech.com/mb/showdoc.aspx?i=3506">http://www.anandtech.com/mb/showdoc.aspx?i=3506 . ;) I care about the best possible performance for the best possible price, with a dash of support thrown in. No hidden agendas around here and if you think so, you would be wrong.
  • yacoub - Tuesday, September 15, 2009 - link

    The rest of us know it's not the case, Gary, but the trolls are just tossing out bait to get you to bite. Ignore 'em.
  • lopri - Wednesday, September 16, 2009 - link

    For real. If the test was done using dual-channel on X58 there'd be a huge outrage from certain users how X58 was arbitrarily disadvantage.

    I have no idea why Gary felt the need to defend himself. Any reasonable reader who followed his articles should know he is the most trust-worthy journalist, bar none. Plus many other writers who know alot about technology but don't know how to write, He is a rare case who actually can write. Look at the first sentense of this article: "We welcomed Anand back into the office with open arms this past weekend." This brought a smile in my face and you can't help but feel disarmed and charmed.

    Dispite it being short, the amount of information in this article is simply overwhelming. What's more, it is apparent that the testings were pursued from actual users' perspectives. Arrogant, nasty, and sensational analysis might help raising controversies (and page views), but in the end what benefits us enthusiasts most is these painstaking and seemingly mundane tests that the author undertakes on behalf of the readers.

    Still, as is often the case, Gary choose to be modest than aggrandizing ("Quick Thoughts" instead of "harder, deeper, baby!" / "choice for me" instead of "Dark knight to rescue the civilization"), and I am very satisfied to find myself identifying with the author as a potential user with regard to the testing methodology. Plus, seemingly impossible multitasking tests on top of myriads of the configurations tested substantiate the author's subjective opinion significantly, and I for one am very grateful that someone else took all that beating!

    Now if I could make a request that could possibly make the author's life even more miserable - any chance to add virtualization to the multitasking mix? With the introduction of XP mode in Windows 7, I think the time has finally come that virtualization has some relevance to desktop users.
  • TA152H - Tuesday, September 15, 2009 - link

    Gary,

    There are a couple of missing data points that would be helpful, and you probably should include with every article regarding benchmarks.

    What were the uncore rates on these processors? Were the i7 920 and 860 running with the same uncore speed?

    Also, were you using the D0 i7 920, which offers slightly higher performance, lower power use, and better clock speed potential, or the older C0?

    Lastly, do you think it was fair for the Lynnfields to have more memory? I'm guessing this might have hurt their results a little in most applications since typically one has to relax timings. Did you test for this at all to see if there was additional latency when adding the extra memory? Probably it made no difference, but it could. Also, why not test the Bloomfield in dual channel memory mode? Certainly someone will use it this way, although not too many people (say someone has good memory from a C2 system, in pairs, he or she might be inclined to use it 'for now', with an eye towards faster memory in the future).

  • Gary Key - Tuesday, September 15, 2009 - link

    The processor steppings are in the setup table. Yes, we did use the D0 stepping on the 920.

    The uncore and NB speeds are in the CPUZ screenshots in the gallery. The boards were all run at their stock ratios (except the 965BE as explained) at both base speeds and when overclocking.
    Overclocking -
    920 - 3407MHz
    750 - 3210MHz
    860 - 3612MHz
    965BE - 2200MHz

    Memory Timings were kept as equal as possible between all the systems,except for the 790FX/965BE as explained. The only difference in memory timings between the Bloomfield and Lynnfield systems was tRFC, 88 for Bloomfield, 89 for Lynnfield.

    Running 8GB of memory did not hurt the Lynnfield systems, in fact, a few of the numbers were actually better, mainly WinRAR and AutoCad that will be shown in the full review. I have a lot of motherboards to review and cannot spend any further time comparing additional items like dual channel on 920 vs 860. I will see if we can do that in a memory specific article.

    Right now, we ran the numbers based on stock ratios, adjusted the base timings to ensure equality, and did our best to properly tune each system to ensure the settings were as close as possible. Even the Bclk speeds were changed to get as close as possible between the Intel systems, although we were still off 4MHz.

    We can play the "what if" game all week long, what if the Bloomfield Uncore was increased to 3800, what if CAS6 timings were used on each system, what if the 965BE NB speed was at 2800MHz, what if we allowed the P55 setup to run memory at 2400MHz instead of curtailing it to 1600 to match the 920/965BE, and the list goes on. Guess what, none those fundamentally change the results between each platform.

    Anand will have additional results in the near future, in the meantime, I have a lot of retesting to complete with different video cards for the P55 roundup. :)
  • TA152H - Wednesday, September 16, 2009 - link

    Thanks for the information, I guess I should look closer at the pictures.

    So, basically, what you're saying is, the i7 920 generally outperformed the 860, despite using a considerably slower uncore, which lowers performance for the 920 not only in memory access, but also lowers the level three cache.

    Hmmmmm, and since it's actually easier to run the 920 at higher uncore speeds, based on the flexibility of it, it's a little strange you didn't think this was worth mentioning except to bury it in a picture.

    Really strange, indeed.

    It just verifies my conviction that you guys are doing everything you can to make the P55 platform look good, although you've traded some of your reputation doing so. I'm not suggesting you're getting any money or any stupid stuff like that, just that for some reason you guys fell in love with the P55, and now give 'information' to further that claim.

    Not mentioning the uncore is much faster, on a comparison that was supposed to be apple to apple, is a SERIOUS oversight.

    Clearly, the 920 would outperform the 860 at the overlcocks at the same uncore. But, you wouldn't want to show that, right?

    The i7 920 is a real tough act to follow. If Intel kills it, it would greatly expand the actual usefulness of the Lynnfield. That's why I fear they might. Then again, they might just leave it for the technical savvy since the number of these users is relatively small and the Lynnfield should sell well to the hoi polloi.

  • pervisanathema - Tuesday, September 15, 2009 - link

    There are some really good trolls around here. I'm impressed. Thanks for the laughs.
  • gstrickler - Tuesday, September 15, 2009 - link

    After reading the article and the updates (great job Gary), here's the conclusions I draw:

    1. You can't go wrong with any of these CPUs or motherboards. They're all very fast, your budget and computing needs are more important than the mostly small performance differences between these systems.

    2. If you want to OC or want lower power/heat, go with an Intel Core i5/i7 CPU.

    3. If you're a hardcore gamer, use an AMD/ATI Radeon card on Core i5/i7 based systems or an Nvidia card on AMD based systems.

    4. You don't have to spend more than $300 for a quad core CPU that will deliver top notch performance.

    5. The Core i7/860 is comparable to the Core i7/920 in nearly every respect.
  • crimson117 - Tuesday, September 15, 2009 - link

    "It is items like this that make you lose hair and delay articles. Neither of which I can afford to have happen."

    Thank you for making me almost choke on the scone I was eating.
  • Ocire - Tuesday, September 15, 2009 - link

    That's some strange numbers you got there...
    Just an idea that popped up my mind: Could effective PCIe bandwidth be the key here?
    You can do the bandwidth test that comes with nVIDIAs CUDA with the shmoo-option for pinned and unpinned memory on both platforms.
    If you get higher numbers on the AMD platform, it could be that with P55 and X58 the card is in some cases interface bandwidth bound. (Which isn't that uncommon in some GPGPU applications, too)
  • Holly - Tuesday, September 15, 2009 - link

    I was thinking about bandwidth as well, but then I realised with P55 CPU has direct lanes to first PCIe 16x slot while X58 platform runs via Northbridge... It's way too different approach to both produce same problem imho...

    This seems like driver issue to me. Maybe the CPU and GPU parts of the engine+drivers run asynchronous and communication in between gets suffocated (st like i7 manages to compute too fast and data have to wait for next loop to go through)...

    no clue though, it's just my best guess...
  • TA152H - Tuesday, September 15, 2009 - link

    A lot of people are getting confused by the PCIe being on the processor as opposed to the Northbridge. Because the IMC shows advantages on the processor die in terms of memory latency, and moving the floating point unit did with the 486, it's assumed moving things onto the processor die should be faster. If you look closer though, it becomes clear this just isn't so.

    The reason the memory controller on the processor has advantages comes down to two important considerations - it's transferring to and from the processor, and it's got finer granularity than something running at lower clock speeds.

    Let's look at the second first. Let's say I'm running the processor at eight times the speed of the northbridge. The number isn't so important, and it doesn't even have to be an integer, to illustrate the point, so I'm just picking one out of the air.

    Let's call clock cycle eight the one where the northbridge also gets a cycle and can work on the request from the processor, and for the sake of simplicity, the memory controller can work on it. If I get something on processor cycles one through seven, I could start the memory read on the IMC, but the slower clock speed doesn't that level of granularity, so the read, or write, request waits. This is a gross oversimplification, but you probably get the point.

    Perhaps more importantly, you're transferring from memory, to the processor. It's too the actual device the memory controller is attached to. And since the memory controller is on the processor itself, there's less overhead in getting generating the request to the memory controller.

    The PCIe 16 slot is a different animal. You're not generally using the processor for this, except for now. It's going from one device, typically to memory, with possible disastrous consequences for the brain-damaged Lynnfield line.

    With a proper setup, there wouldn't be any real difference. I'm not sure how the Lynnfield is set up though, and I'm not sure it's a proper setup like the x58. I'll explain. If we look at the x58, the video card will transfer memory to the Northbridge, and then the Northbridge will interface to the memory. It has channels on it to do that, and it doesn't involve the processor at all, or need to.

    On the Lynnfield, now you're involving the processor, for no good reason except for cost (which is a really good reason, actually). So, now the video card sends a request to the PCIe section of the processor, and the processor has to do the transfer. Now, this is the big question - just how brain-damaged is the Lynnfield? Does it actually have a seperate path to handle these transfers, or does it basically multiplex the existing narrow memory bus to handle these? It's almost certainly the latter, since I don't know how much sense it would make to only use part of the memory bus for all other transfers. Anand can hopefully answer these questions, although this site tends to never look very deep at things like this, so I'm doubtful. I guess they don't think we're interested, but I surely am.

    The end result is, you could have competition for the already narrower memory bus, where the processor can get locked out of it while PCIe transfers are going on. This is consistent with some benchmarks other sites show, where the Lynnfield struggles more than it should on games.

    I don't want to make this sound worse than it is though. Most video cards have a lot of memory, and the actual number of requests to memory hopefully isn't very high. Even on a Bloomfield, any request to main memory is a slow down since video memory will always be faster. Also, cards will keep getting more memory, whereas the human eye is not going to get able to discern better resolution, so presumably cards will not have to use main memory at all, in the future. And, keep in mind, processors have big L3 caches, so don't need to go out to memory all that often. So, it's not catastrophic, but you should see it, and more as you stress the CPU and GPU. Again, if you look at other websites that did some serious game stressing, you do see the Bloomfields distance themselves from their brain-damaged siblings as you stress the system more.
  • yacoub - Tuesday, September 15, 2009 - link

    I think we'd all be interested to know the answer to this quandary, however I also think you're exaggerating a bit when you call it brain damaged and say things like "disastrous consequences". In reality what you mean is that in extreme bandwidth saturation situations like SLI it's possible it might be slower than X58, and maybe when the DX11 generation of cards come out they will actually require enough bandwidth for it to be even more noticeable in SLI comparisons. But for the vast majority of us who run a single GPU, so long as a single DX11 card doesn't fully saturate the available bus bandwidth and thus doesn't perform any less than on X58, P55 is just fine.
  • TA152H - Wednesday, September 16, 2009 - link

    I agree, when I was typing it, I meant disastrous consequences with respect to that particular instance when the processor has to multiplex. In the context of overall performance, I don't think it would be that huge. I wasn't clear about that.

    But, you're off with regards to the saturation. You don't have to saturate the bus, at all. You don't need two cards. That type of thinking is fallacious, in that it assumes only part of the bus is used.

    In reality, ANY time the video card needs the memory bus, and the processor needs to read memory, you've got a collision, and one or the other has to wait. This would happen more often if you have more stress on the processors, or video subcomponent, but could happen even with one card, and one processor being used. It just would be much less frequent, and probably insignificant.

    These are the type of compromises these web sites should be bringing up, but they simply don't. It's not just this one, but shouldn't a tech site bring up questions like this, instead of just publish benchmarks (which, as we know, are far from objective and can paint a different picture based on the parameters and benchmarks chosen).

    I wouldn't expect this from PC Magazine, but from 'tech' sites? They just aren't very technical.
  • Gary Key - Tuesday, September 15, 2009 - link

    P55 will be just fine with the DX11 cards.. whistles and grins evilly looking at the results, however, for top performance in SLI or CF, it is X58 all the way. That said, I have not been able to tell the difference between 130 FPS and 128 FPS in HAWX yet, nor between 211 FPS and 208 FPS in L4D. :)
  • jonup - Wednesday, September 16, 2009 - link

    does your comment mean that you have/have seen a HD 5870 in action?
  • yacoub - Tuesday, September 15, 2009 - link

    It appears more driver/optimization-related given that some games that are less bandwidth-intensive are showing the strange performance and others are not.
  • yacoub - Tuesday, September 15, 2009 - link

    Hey I wonder if Gary tried renaming the .exe just to see if it was a driver bug with certain game engine optimizations! :)
  • Gary Key - Tuesday, September 15, 2009 - link

    I tried renaming the exe files, no go, anyway, testing with the 4890s now and will be finished shortly so we can move on.
  • Ocire - Tuesday, September 15, 2009 - link

    Do you have a link to a recent article showing the PCIe-bandwidth usage of different games? The most recent article doing something related I could find in a quick search was some test over at Tom's hardware testing x1900xtx vs. 8800gts on different bandwidths. ;-)
    Renaming the .exe would be fun though to see whether some strange optimization causes this effect.
  • Eeqmcsq - Tuesday, September 15, 2009 - link

    Thanks for the update and including Turbo off numbers for the i5 750. So at its baseline, the i5 750 is about the same or worse than the Phenom II 965. But the Turbo mode either brings the comparison closer or tips the scales in i5's favor in individual tests.

    As for the multitasking tests, I can understand why the i5 750, even w/Turbo, fell behind the 965. The i5 was already behind in the Lightwave 3D and Cinema 4D individual benchmarks. Combining them together would certainly "worsen" the difference, as shown by the multitasking charts.

    Also, looking at the Cinema 4D test, I noticed how the Turbo boost was much less of a boost during the multitasking test compared to the individual test. This is the predicted effect as heavy multitasking pushes the CPU TDP up, leaving less room for Turbo. As a result, the i5 Turbo performance approaches that of its baseline, which is generally equal to or worse than the Phenom II 965.

    Individual test, i5, lower % means better Turbo Boost effect:
    Cinema 4D - Turbo finished in 84.5% of the baseline time
    Lightwave 3D - Turbo finished in 96.5% of the baseline time

    Multitasking test, i5:
    Cinema 4D - Turbo finished in 95.3% of the baseline time
    Lightwave 3D - Turbo finished in 95.6% of the baseline time
    Total time - Turbo finished in 95.4% of the baseline time

    So thanks again to Gary for providing the baseline Turbo numbers as well as the multitasking tests. It certainly paints a more complete picture of how to measure Turbo capable CPUs.

    Requests: I'd still like to see 1x, 2x, 3x, 4x concurrent instances of the same single threaded test just to further demonstrate the variableness of Turbo compared to its baseline.
  • Malhandir - Tuesday, September 15, 2009 - link

    One aspect of the new 1566 i5/i7 processors I love is the more intelligent turbo boost, if only because it achieves its goals so well. Have there been any indications of this being introduced into new 1366 processors in the future or are there technical limitations to its introduction keeping it to the 1156 platform? Of course any future 1366 CPUs are, well… future! But I’m interested in the answer nevertheless, since the most obvious tweak that could be done on the 1366 CPU lineup is the introduction of this core management ability.
  • Mastakilla - Tuesday, September 15, 2009 - link

    Great article once more!

    I think it is clear in meanwhile that the 920 offers less value for the money than a 860 in almost all the cases (except extreme benchmarking)

    one use case however isn't covered yet (and that is exactly the use case that Intel said the 920 would be better for):
    How do they compare with a decent OC? (meaning 4Ghz+)
    Do the 920s reach 4Ghz+ easier?

    I am talking about a fully stable 24/7 watercooled setup than (cause you can't decently aircool at that speed except for some short benching)

    cause that is exactly what I have in mind for my next system ;)
  • Ann3x - Tuesday, September 15, 2009 - link

    You need less volts to stabilise 920s than lynnfields (due to the pci-e stabilisation issues) therefore they run a little cooler at load therefore *should* overclock better.

    I say should as the lynnfields havent been around long enough to say 100%.

    The opinion about the 920 being worse vfm is subjective. Atm the 920 is actually cheaper than the 860 in most places, this is offset tho by x58 vs P55. Its a close thing either way. The P55s are certainly a good bet if you dont want to push overclocks (and tweak a lot) and dont want a multi GPU setup. If you do want either of these the 920 is probably better.
  • Alberto - Tuesday, September 15, 2009 - link

    I don't understand the botton line of the article. We can not make a comparison between available cpus and unavailable ones. Phenom II 3.4 Ghz is an exotic and hot cpu, moreover it's pretty rare to find in stores, same thing for the 3.2Ghz incarnation.
    A real apple to apple comparison would be between Phenom II X4 2.8/3Ghz and i5/720...both 95W and available cpus.
    Gary, you can't say "in multitasking my chose is......". The real story is that you can't find a Phenom II X4 965/955 Black configured machine from a good OEM. These are exotic cpus for expensive gaming systems and AMD isn't able to make many of them.
    Core i7 is everywere on big OEM sites.
  • Makaveli - Tuesday, September 15, 2009 - link

    Alberto Who the Fu*k buys OEM systems on this site?

    I can find both those cpu's at any local stores in my area.

    Please put the pipe down!
  • Alberto - Wednesday, September 16, 2009 - link

    Anyway these Amd cpus are marginal and rare around the world. They don't make good revenue for a Red company like Amd. Maybe you don't have a real job, without the on site assistance of a big OEM i'd be ruined.
  • coconutboy - Tuesday, September 15, 2009 - link

    Anandtech tends to be an enthusiast site where (I assume) a heavy percentage of the readership builds their own computers by handpicking their parts. This means OEM machines are not the main focus. As for your claim that some of those AMD cpus are not easy to find, perhaps you do not live in the USA. Otherwise, I see the AMD 965, 955, 720 etc for sale at every one of the 4 retailers I just checked.

    newegg
    mwave (I can walk into this store locally for purchases)
    microcenter (another walk-in store)
    zipzoomfly

    I'll bet Fry's and bucketloads of other walk-in retailers offer them as well. Now we just need AMD to drop the prices to compete more evenly with Intel (although the x3 720 is quite inexpensive)
  • GeorgeH - Tuesday, September 15, 2009 - link

    *Puts on tinfoil hat*

    Intel won't let Nvidia make chipsets for 1156/1366.

    Nvidia GPU's perform conspicuously poorly only on 1156/1366.

    Coincidence???? You decide!!!!!!!

    *Takes off tinfoil hat*


    Random question: Does the Nvidia+Intel performance thing correlate at all to how multithreaded a game is?
  • CB434 - Tuesday, September 15, 2009 - link

    You know it's an honest review when unexpected results like this pop up.

    Purely coincidentally, I have emailed a few reviewers in the last few weeks about this..

    All the video card reviews for the last 12 months show i7's as the test rig for all the video cards. It's so stupid that no one has ever thought to use an Nvidia card on a Phenom. Until now! Good work. It's always either AMD + Ati or I7 + Ati or Nvidia. All of the perceptions and ideas of performance for the current gen of Nvidia video cards is all based on how they work on an INTEL. Not overall.

    It's just a shame the Phenom doesn't have SLI/CF in the same board. I'd be using Phenom II with 275GTX SLI and looks like it would be a kickass solution for gaming.

    I don't think it's necessarily a bug or a problem. Maybe it's just the lottery/fluke of when you combine different parts that have different ways of working. Ati and Nvidia go two seperate ways to reach the same goal. Different memory bus bandwidth, shaders clocks etc and maybe something in there gels with the way the Phenom II works.
  • FlanK3r - Tuesday, September 15, 2009 - link

    AMD X4 965 with 2200 MHz uncore

    2200 MHz uncore is whorse than default 2000 MHz! Its one bad for review :(...I wrote review about Phenoms overcloking and made some comparsion uncores vs CPU cloks. In all performance, 2200 MHz uncore is very bad choice. More better is 2000MHz and of course 2400 Mhz or 2600 MHz (and higher for CPU clocks 3700 MHz up)
  • daydr3am3r - Tuesday, September 15, 2009 - link

    I have to ask, and pardon me for the triviality and/or ignorance but,
    why is the article title

    Topic: Motherboard
    Manufacturer: ASRock

    ?

    The accompanying picture also displays an ASRock mobo..
  • Ben90 - Tuesday, September 15, 2009 - link

    uhh, so AMD pretty much wiped the floor BIG time with the gtx 275. Ive heard several reports that AMD chips perform better on higher resolutions then i7s, but after research i never found the results to be that drastic

    I hope to hear the answer, maybe Nvidia released a driver that absolutely loves the deneb architecture, or possibly AMD just got a lucky two games as the ones benched.

    If its something like the first one, and phenom can do this consistently game to game, we might see some big changes soon.
  • MadMan007 - Tuesday, September 15, 2009 - link

    Thanks for the exploration of overclocks but unfortunately this article is worthless to me, and perhaps many others, without any Core 2 CPUs. Unless I want to try to extrapolate back to older articles, and that's a guessing game plus I'd need to find overclocked results, I am unable to tell how much benefit a new system would be.
  • ggathagan - Wednesday, September 16, 2009 - link

    As Gary clearly stated on the first page of the article and reiterated in the third comment, he had not yet finished with the P45/Q9550 testing at the time of his update.
    He also clearly stated that he would be adding those results at a later point.
  • coconutboy - Tuesday, September 15, 2009 - link

    Good update Gary, this is the kinda article/update I like to see. Comparisons are tough, but using dollar-for-dollar or clockrates helps me as a consumer.

    I notice that for walk-in customers, Microcenter (at least the one near me here in SoCal) has the following prices

    $230 i7 860
    $200 i7 920
    $160 i5 750

    I can get a 920 for less than an 860, then combine it with the ASRock X58 extreme that was recommended back in the July article ($170 at Newegg and it is getting high marks in the comments) and do a moderate overclock to 3.2-3.8GHz to achieve amazing performance for the price. Alternatively I can pick up the i5 750 and a ~$100-130 mobo and have a very low-cost outlay for a great gaming box. Hopefully AMD will also drop their CPU prices soon to give us yet another option.

    With the new AMD and Nvidia cards coming out in the near future, all these choices are very inexpensive for the performance and will save $$ to be spent on a brand-spanking new vid card. In the meantime, an ATI 4850 or Nvidia 9600GT can be had for under $100 to conserve $$ and tide gamers over till the new hardware hits. I'm building two new systems in the next few weeks, and the above is my gameplan.
  • MadMan007 - Tuesday, September 15, 2009 - link

    Anyone who is a gamer/enthusiast who doesn't have at *least* an HD4850 or 9600GT shouldn't be buying either one right now. Only if they're building a complete new system and are too desparate to wait.
  • cactusdog - Tuesday, September 15, 2009 - link

    Not worth running 3.8Ghz if your temps are 90 degrees with the best air cooler money can buy. The 1366/1156 are great but we are hitting a temp ceiling now. People should be made aware of this because they will buy it, take it home and realise they cant really run at those settings even with the best air cooler. If you cant run at those settings the gap widens between the 920.

    This issue is being completely ignored or glossed over.
  • yacoub - Tuesday, September 15, 2009 - link

    We really need the P55 motherboard reviews. That's where we (hopefully) learn about temps, stability, power draw, features, etc. I certainly won't be buying until I read reviews for the P55 boards.
  • Stradigos - Tuesday, September 15, 2009 - link

    I agree! I can't wait to read it. I won't be buying anything till I do.
  • strikeback03 - Tuesday, September 15, 2009 - link

    I'm waiting on the same thing. The ~$50 difference between a P55/860 system and an X58/920 system isn't enough to affect my decision much, but this article mentions in passing that the P55/860 system uses 70W less, which would be a definite consideration for me. I'm waiting to see whether this is only in stock configuration or if there is also a difference when overclocked, temps on good air coolers, etc.
  • jonup - Tuesday, September 15, 2009 - link

    I'm with you. I am upgrading within the next 6 months and if the i5/p55 is worth the money I will be jumping on it sooner rather than later. Microcenter has the i5 for $170 after MI sales tax. I would of pull the triger but there is not enough MB reviews yet.
  • Gary Key - Tuesday, September 15, 2009 - link

    I am not going to wait much longer, just about every person I have contacts with is working on the NV problem. I just hate showing numbers with the knowledge that something is not right. It is not just games, the 3D Rendering apps like Cinema 4D are affected also. That said, I have spent the better part of the last three days testing with the HD4890. ;)
  • CB434 - Tuesday, September 15, 2009 - link

    Gary, this GPU issue...

    Is it that the i5 and i7 is "underperforming" and is getting less FPS than it is expected to (relative to other benchmarks). Or is it simply that the AMD/275 combo is simply better then expected? Is there a "problem" with the i5/i7 and the 275GTX per se? If there was no AMD/275 test, would you have thought there was a problem?

    I'm confused about which is which.

    Are your contacts Intel people, trying to find a reason for the "problem" or are they AMD people trying to find an explanation for their "success"?

    For the record I've been hunting down 275GTX reviews for quite a while and this is the first I've been able to find using a Phenom II and a 275GTX (no chance of ever finding 275GTX SLI on Phenom :(.. ) It's always an i7 in the GPU test rigs. So finding a frame of reference or baseline for how 275GTX is supposed to work with Phenom CPU's is hard because there is nothing out there.
  • GeorgeH - Tuesday, September 15, 2009 - link

    From other reports I've seen on this issue, it really is a problem and it's specific to 1136/1366 - a high-end 775 C2Q would have performed pretty much identically to the X4 965.

    Bottom line, LGA775+GT200 looks to be a superior gaming platform to a i7+GT200 right now.

    Source (bottom of page):
    http://techreport.com/articles.x/17545/6">http://techreport.com/articles.x/17545/6
  • yacoub - Tuesday, September 15, 2009 - link

    So is this issue unique to the P55 boards, or do the X58 (or whatever the i7 920 runs on) exhibit this same issue when running the same NVidia drivers and Win7?
    If the issue isn't unique to P55, it's not really a factor in which P55 motherboard to buy if they all exhibit the same behavior and the X58 boards do as well.
  • CB434 - Tuesday, September 15, 2009 - link

    That's why turbo mode is so cool.

    You can get all the upsides of a 24/7 3.8Ghz overclock without the extra heat and power usage. You can OC to 3.2 24/7 and let turbo do the rest. For single and dual core turbo it will shut off the other 2/3 cores, meaning more hz for less power consumed and heat generated. Turbo can give you all the upsides without the downsides, or at least the downsides are reduced.
  • jeffj29 - Tuesday, September 15, 2009 - link

    So it looks like OC'd at 3.8 GHz with HT on that it's a dead heat performance-wise between the x58/920 and then p55/860 platform.

    The 860 combo would have lower cost but the 920 combo have greater future upgrade performance potential (with a future 6-core processor). But here is the question I am most curious about... how much higher would a 920 (with D0 stepping) overclock to using air cooling than a 860? For example, if the 920 overclocks easily to 4.0 GHz but the 860 wouldn't OC stable past 3.8 GHz then the 960 would be the clear performance pick in my book.
  • MadMan007 - Tuesday, September 15, 2009 - link

    I don't see many enthusiasts, even upgrade junkies, buying a $1000-1500 6c/12t CPU. It will likely only find a home in time=money systems as far as workstations.
  • nvmarino - Tuesday, September 15, 2009 - link

    I remember reading about an issue with nvida performance on i7 back around i7 launch. Yep, here it is (one of the few decent articles I had read over at tom's in a while...):
    http://www.tomshardware.com/reviews/geforce-gtx-28...">http://www.tomshardware.com/reviews/geforce-gtx-28...
  • Ryun - Tuesday, September 15, 2009 - link

    I was just about to post this as well. I've seen other sites come across the same issue as well: http://www.bit-tech.net/hardware/cpus/2009/04/23/a...">http://www.bit-tech.net/hardware/cpus/2...enom-ii-... (Far Cry 2 benchmark is more noticeable)

    I would have thought the issue would've been fixed by now, though I still have not recommended GTX 2-series cards wtih Nehalem in spite of it.
  • Gary Key - Tuesday, September 15, 2009 - link

    Here is thing, the numbers lined up with the AMD HD 4890 the last time I tested on Vista with the 180 series drivers. I tried the 180s under Win7 and had the same problem, even the inbox Win7 drivers show this pattern.

    This does not occur in all games either, which makes it even more confusing as the thought process a few months ago was the lack of driver optimizations in GPU bound situations. If that were the case, Crysis/Crysis Warhead should show the largest difference based on current game engines, it does not.

    Instead we have titles like H.A.W.X. and L4D, not exactly GPU killers, showing this pattern besides FarCry 2. We just want an answer, but I am not going to wait much longer. ;)
  • neoflux - Tuesday, September 15, 2009 - link

    Excuse my noobness, but are those 2 gaming benchmarks in frames per second, meaning that higher is better? Also, what is the minutes for each processor for? Thanks.
  • Rajinder Gill - Tuesday, September 15, 2009 - link

    Higher is better. 'min' refers to minimum frame rate.
  • yacoub - Tuesday, September 15, 2009 - link

    yeah i was wondering why "minutes" were listed. might want to change "min" to "min fps" or something similar. "min" more commonly means "minute" and leads to confusion.
  • neoflux - Tuesday, September 15, 2009 - link

    thank you, sir.
  • SmCaudata - Tuesday, September 15, 2009 - link

    I think with the current generation of video cards the 1156 is the way to go. The question I have is what about the next gen. With all of the games that are now video card limited it seems that the cheapest upgrade to a new computer down the road will be a second vid card. It will be interesting to see how much the 1156 bottlenecks with CrossFire in the next few weeks.
  • yacoub - Tuesday, September 15, 2009 - link

    Most people aren't worried about multi-card systems. So long as the single card next-gen GPUs don't fully saturate a PCIe x16 bus, 1156 will be fine.
  • MODEL3 - Tuesday, September 15, 2009 - link

    Nice update Gary.

    We wondering in the forums about Windows 7 HT on/off performance.

    Do you think that this kind of performance difference is indicative?

    Also, from your experience so far with Windows 7, how much do you think Windows 7 improves the HT performance in relation with Vista?


    Thanks,

    MODel3
  • Gary Key - Tuesday, September 15, 2009 - link

    From an objective viewpoint, it is hard to quantify what the measurable improvements are with the revamped scheduler in Win7, mainly due to all of the other improvements. Subjectively, that means watching the actual activity in the performance monitor, the core loads are certainly balanced better now. Also, in the multitask test, Vista 64 would have choke at times. One or more of the apps would actually stop/stutter during testing before continuing on. I have not seen that in Win7 so far. ;)
  • Patrick Wolf - Tuesday, September 15, 2009 - link

    Why no Core 2's included in the benchies?
  • elivebuy - Tuesday, December 8, 2009 - link

    http://www.elivebuy.com">http://www.elivebuy.com

    Air jordan(1-24)shoes $33

    UGG BOOT $50

    Jewerly $20

    Nike shox(R4,NZ,OZ,TL1,TL2,TL3) $35
    Handbags(Coach lv fendi d&g) $35
    Tshirts (Polo ,ed hardy,lacoste) $16

    Jean(True Religion,ed hardy,coogi) $30
    Sunglasses(Oakey,coach,gucci,Armaini) $16
    New era cap $15

    Bikini (Ed hardy,polo) $25

    FREE sHIPPING
    http://www.elivebuy.com">http://www.elivebuy.com
  • Kaleid - Wednesday, September 16, 2009 - link

    I'd love to see how c2d's manage when o/c to 3.6-4Ghz range.
  • Gary Key - Tuesday, September 15, 2009 - link

    I was not finished with the P45/Q9550 testing yet. Anand will have those numbers in his article and I will update my results once I have completed testing.
  • jonup - Tuesday, September 15, 2009 - link

    I agree. You guys put up with the crazy demands of snakeoil to disable standard features, but will not include a C2Q! Doesn't C2Q overclock 200-300MHz above anything on the market? Not to mention that C2Q run cooler and clock for clock perform equal to PII.
    Please, include C2Q and max 24/7 stabe OC numbers in the full review.

Log in

Don't have an account? Sign up now