Comments Locked

150 Comments

Back to Article

  • damage98 - Saturday, November 21, 2009 - link

    I have an asus m4n78 pro mobo. Would the new gt240 be appropriate?
    Thanks!
  • Archer0915 - Friday, September 25, 2009 - link

    Well I have put it through the ringers and this is what I came up
    with: http://www.techreaction.net/2009/09/25/athlon-ii-x...">http://www.techreaction.net/2009/09/25/...-x4-620-...

    This thing can smoke or at least keep up with the common PhII or Core 2
  • monkeyman1140 - Monday, September 21, 2009 - link

    I'm kinda iffy about how this compatibility thing works, and it seems manufacturers aren't terribly interested in compatibility bios updates either, preferring you to fork over fresh cash for the latest mobos.
    I'd like to put this in an older dual core system thats perfectly fine but its just not as fast as it used to be...
  • flexy - Friday, September 18, 2009 - link

    What version of Cinebench R10 are you using?

    The 64 bit version or the 32 bit version?

    G.
  • ClagMaster - Thursday, September 17, 2009 - link

    Another well written article by Mr Shimpi on the latest AMD mainstream quad core offering. Article was brief and to the point with adequate benchmarking to support his claims. It's articles like this that keeps me coming back.

    The i5/P55 is the mainstream processor par excellance to acquire for a major upgrade if you presently have an Intel rig. This is what I am going to upgrade to next year because I have a 3 year old Intel rig. By that time, there will be 65W Lynnfields available with P55 boards thoroughly debugged.

    However, if you have an AM2+ motherboard in good shape with a 780G/785G/790X/790GX chipset with continuing BIOS support, then the Athlon II X4 620 is an outstanding upgrade from dual to quad core for $100-$120. This is a really good value for bargain or mainstream. Price:Performance ratio is better than i5 just on the basis of the CPU alone. Throw in a paid-for motherboard into the equation and it gets even sweeter.
  • jtleon - Thursday, September 17, 2009 - link

    Great Article Anand - as usual!

    Despite the other Intel Fanboy comments here, I take away from this that AMD is bringing Quad to the masses - and undercutting the competition at the same time.

    Running an old Athlon XP as I write this, I am glad to see AMD resurrecting the Athlon name, and applying it to what may be their new bread & butter piece of silicon.

    Clearly in a depressed worldwide economy, performance takes a backseat to price - AMD has an ace here with this design, in its 1st iteration, appears to have Intel over a barrel with regard to their inflated price structure. From the benchies here, the performance differences are almost imperceptible. Thus the Athlon II based boxes should jump off the shelves, leaving the other guys gathering dust.

    Kudos to AMD - and Best of Luck on the next gen Propus.

    jtleon
  • Genx87 - Thursday, September 17, 2009 - link

    That X4 955 is being smoked for the most part by the i5 750. Intels basement i series processor. The i5 performs better, costs less, and consumes less power.

    Why cant AMD get their act together? Ever since Core 2 Duo they have been on the wrong end in a bad way.
  • the zorro - Thursday, September 17, 2009 - link

    that's false.

    the results are biased because core i5 750 its overclocked at least 600 mhz.
    phenom 955 beats core i5 750 clock by clock.
    also when overclocked to 4 ghz core i5 temperatures are almost 100 C which is a failure.
    also core i5 power consumption when overclocked skyrockets because of the integrated northdbridge.
  • Genx87 - Monday, September 21, 2009 - link

    That is really irrelevant to the avg user is it not? The avg user doesnt care how the processor achieves it power\performance. Only that it does. That is a design feature of the Intel chips that isnt in AMD. Bottom line is in the suites and everyday use AMDs top processor is often beat by Intels next gen entry level chip.
  • silverblue - Thursday, September 17, 2009 - link

    Clock for clock? I don't think so. True, if the i5 didn't have Turbo, it wouldn't sprint ahead so far in single threaded applications, but the fact is it does and it's a legitimate technology. However, the 955 pulls closer, clock for clock, in multi-threaded tasks.

    The i5 ships with a rather weak cooler. It's not suitable for heavy overclocking... but then again, if you want to do it right, you'd get an after-market cooler anyway.

    Nothing that AMD has out now is better clock-for-clock than Core2 or Nehalem, no matter how much we'd like to believe there is.
  • the zorro - Thursday, September 17, 2009 - link

    the turbo overclocking is plain overclocking of all lynnfield cores at least 600 mhz and you are comparing lynfield overcloded results versus phenom 2 955 stock speeds.

    phenom 2 is much better than lynnfied 750 and when overclocked to 4ghz remains at 55 C while lynnfield temps are almost 100C. which sucks.

    all the 'advantage' of lynnfied in these results comes from benchmarking an overclocked processor and present it as if it were stock speed, which is illegal
  • silverblue - Thursday, September 17, 2009 - link

    Wrong. I'll help you here...

    i5 750
    Default: 2.66GHz
    4 cores: 2.93GHz 3 cores: 2.93GHz 2 cores: 3.06GHz 1 core: 3.20GHz (gains: 266MHz/266MHz/400MHz/533MHz)

    i7 860
    Default: 2.80GHz
    4 cores: 3.06GHz 3 cores: 3.06GHz 2 cores: 3.33GHz 1 core: 3.46GHz (gains: 266MHz/266MHz/533MHz/667MHz)

    (all data reassembled from the second table at http://www.anandtech.com/cpuchipsets/showdoc.aspx?...">http://www.anandtech.com/cpuchipsets/showdoc.aspx?...

    The turbo mode gives a minimum boost of 266MHz and a maximum of either 533MHz for the i5 or 667MHz for the i7. NOT 600MHz for all cores. Quite where you got that from is beyond me. Additionally, Turbo is a 100% legitimate technology. Would you be happier if the default clock was throttled?

    As for illegal... we have a word in the UK for this sort of comments, and it's "bollocks".
  • Eeqmcsq - Thursday, September 17, 2009 - link

    Good chart, though you linked to the wrong page. It was on the Turbo Mode page. I have no idea how I missed that chart the first time I read the article.

    But for the i5 750, if the lowest GHz Turbo mode will operate at is 2.93 GHz, why doesn't Intel just call it an i5 750 at 2.93 GHz, instead of 2.66 GHz?
  • silverblue - Thursday, September 17, 2009 - link

    My bad - the link has a ) after it which sends you to the intro page. I should've put a space after the link.

    At least you checked through to find it... I doubt "the zorro" has bothered as of yet.

    You make a very good point about Turbo but it's not as if Intel is pretending it doesn't exist or it's disabled; it should be relatively easy for people to make comparisons using the benchmarks in this article as well as others. I suppose it may have helped if there was data suggesting the clock speed at the time of a specific test, but I can't imagine that'd be a very easy thing to do especially if threads are being bounced across cores and as a result, the clock speed is constantly fluctuating. The alternative would've been to disable Turbo and that would've prevented people like "the zorro" from his pointless tirade on here, but they would've been disabling a feature of the processor to test it against rivals that lack that feature.
  • Eeqmcsq - Thursday, September 17, 2009 - link

    > I suppose it may have helped if there was data suggesting the clock speed at the time of a specific test, but I can't imagine that'd be a very easy thing to do especially if threads are being bounced across cores and as a result, the clock speed is constantly fluctuating.
    - No, but the troll does make one point that I agree with. When an i5 is run with Turbo mode, Anand's charts should NOT list that it is at 2.66 GHz. He should list it as 2.93-3.20 GHz, especially if it is 100% certain that Turbo mode NEVER reaches the baseline Turbo off clock of 2.66 GHz.

    > The alternative would've been to disable Turbo and that would've prevented people like "the zorro" from his pointless tirade on here, but they would've been disabling a feature of the processor to test it against rivals that lack that feature.
    - Well, I've asked for the same thing, but for different reasons that I don't want to repeat in this comment or I'll sound like a troll. :)
  • the zorro - Thursday, September 17, 2009 - link

    if you are an accountant and do the same thing and try to present false results as real then you would go to jail.
    you can call a bank robbery an 'auto loan' but still is robbery and you still go to jail.
    people is not stupid.
  • silverblue - Thursday, September 17, 2009 - link

    AMD, please just incorporate a turbo mode into your next CPUs so we can get rid of trolls like this one. Do it for me.

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

    no amd wont do that.
    why?
    because a platform at stock speed is more stable than other that is auto overclocking, also intel created the turbo crap story to charge for the overclocking.
    yes intel is charging their users for the overclocking,now overclocking is a feature and intel charges for it.
    amd phenom 2 overclocking is free.
    try to auto overclock a server, that is not good and creates instability in the platform.
  • silverblue - Thursday, September 17, 2009 - link

    And if they do, oh what will you do then?

    Additionally, try to explain to me what the 965 BE is other than an overclocked variant of the 955, which in turn was an overclocked variant of the 945. Isn't it interesting how none of these overclocks better than the others, and why the 965 BE has a higher TDP than the 955?

    You act as if AMD and Intel have never produced an overclocked variant of any of their CPUs until Nehalem turned up.
  • the zorro - Thursday, September 17, 2009 - link

    i can see that you have no clue, when a new processor is introduced that is 200 mhz faster than another model, is not an overclocked processor, is a more refined silicon, with better electrical properties and more stable at highers speeds. and also with higher overclocking headroom.
  • silverblue - Thursday, September 17, 2009 - link

    Except for you I suppose?

    Imagine you own a car with a turbocharger. When you accelerate, the turbo just happens to kick in. Is it illegal for the manufacturer to have put the turbo in there in the first place?

    I agree it's not a completely parallel analogy due to the fact that a turbocharger provides a free power boost from otherwise wasted energy, but the point is that it's a part of the design and people have, and always will, accept that.

    How are the Lynnfield's results false? It clearly states in the nomenclature that the Lynnfield possesses a turbo mode designed so that the processor operates at a higher clock speed for a stated number of cores dependent on the load being imposed upon it. Imagine if a new SIMD instruction set appears and only AMD processors can use it, and software is accelerated as a result - is this illegal?

    I feel my intelligence slipping away as I try to reason with you.
  • the zorro - Thursday, September 17, 2009 - link

    turbo overclocking is just overclocking, nothing else.
    overclocked results are illegal if you try to present them as stock speed results.
    you can call overclocking 'banana' if you want, but still that banana is illegal
  • Kaleid - Saturday, September 19, 2009 - link

    There is nothing illegal about it. Repeating it doesn't make it so.
  • silverblue - Thursday, September 17, 2009 - link

    But they haven't! Why can't you see this?

    Let me put it another way. In single core mode, with turbo enabled, the i5 750 is clocked at 3.2GHz. The Phenom II X4 965 BE is STILL clocked at 3.4GHz and it STILL loses. And guess what, add turbo to more cores and they all slow down, thus making the 965 BE look even worse in most situations as its clock speed gap increases even further.

    I've read over those 3.8GHz results. No turbo mode (not illegal, then!). i5 still wins most of the benches. Granted, some of the tests are Intel-optimised applications, but a) it's not Intel's fault that AMD optimisations are lacking from specific programs, and b) in games with no specific optimisations for either architecture, the Lynnfield is still going to win because even with turbo enabled, the i5 750 is STILL clocked lower than the 965 BE and it's STILL equalling or beating AMD's strongest CPU.

    End of.
  • the zorro - Thursday, September 17, 2009 - link

    don't fool yourself by over-thinking to justify an illegal activity.

    turbo overclocking is just overclocking.

    and phenom 2 955 beats core i5 750 at stocks speeds, that is lynnfield without overclocking.

    lynnfield is a failure, because when overclocked to 4ghz temps are almost 100C. and power compsumption skycket.

    phenom 2 overclocked to 4ghz is cool at 55C.
  • the zorro - Thursday, September 17, 2009 - link

    if you overclock phenom 2 955 the same 600 mhz that lynnfield is overclocked, it wipes and mops the floor with lynnfield 750.
  • strikeback03 - Thursday, September 17, 2009 - link

    Umm, check your numbers again. Unless you want to claim a 955 at 3.8GHz would perform differently than a 965 at 3.8 GHz. See here: http://www.anandtech.com/mb/showdoc.aspx?i=3639&am...">http://www.anandtech.com/mb/showdoc.aspx?i=3639&am...

    All the 3.8GHz numbers are clock-for-clock, with no turbo (and no HT since the 750 doesn't have it) and the only test where the 965 tops the 750 is the Lightwave3D portion of the multitasking test. The 750 does beat the 965 in the overall multitasking test.
  • the zorro - Thursday, September 17, 2009 - link

    as you can see, at stock speed, when lynnfield 750 has turbo overclocking off, then phenom 2 965 annihilates lynnfield.
    look at all the tests.
  • strikeback03 - Friday, September 18, 2009 - link

    Well, yes, a 965@3.4GHz does beat a 750@2.66GHz, AMD's architecture isn't that far behind in most cases. But the only ones likely to leave the clocks stock are OEMs, who will also leave the turbo mode which seems to be the bane of your existence turned on. In stock configuration for both processors and looking at only the tests Gary conducted in the above article, the 965 carries advantages of 10%, 19%, -8%, -2%, -6%, 7%, and 12% over the 750. Using pricegrabber, Newegg has the best prices for each right now, at $199.99 for the 750 and $245 for the 965BE. This is a 22.5% difference in price, the performance gain is not that high in any of the tests. This is also ignoring all the tests in Anand's i5/i7 launch article, the majority of which the 750 topped the 965 in.
  • mdk77777 - Thursday, September 17, 2009 - link

    You buy a new Computer every five minutes?
    Really, I5 750 started shipping like a few days ago.
    Competitive product, but requires a new MB.

    Declare the end of the war after 1 Second of battle doesn't make much sense.
  • Genx87 - Monday, September 21, 2009 - link

    This losing battle for AMD has been going on since the introduction of the Core 2 Duo years ago. Where have you been?
  • mdk77777 - Wednesday, September 16, 2009 - link

    945 has been at $170 for weeks. That's a $55 difference from the price listed by ANAND :!: , it is also available at 95 watt :!:
    955 has been at or below $200

    But that would conflict with the second coming of I5 at $210 so he conveniently ignores current pricing. :mrgreen:

    Talk about bias, you write an article about pricing and market position and then ignore current pricing and market positions.

    Pretty amazing.
  • Voo - Wednesday, September 16, 2009 - link

    You do understand what the listed prices are, and just want to flame, right?
  • mdk77777 - Wednesday, September 16, 2009 - link

    How does AMD respond to Lynnfield? Is it by drastically cutting prices on Phenom II? Nope.

    Thats how the article starts. Correct, but they lowered prices weeks ago in anticipation.

    A list price means nothing if it has been widely, significantly, and uniformly discounted.

    Obviously they will update. Again, a story about market prices and market value of various CPU should reflect the market pricing.
  • yacoub - Wednesday, September 16, 2009 - link

    you are correct that the actual prices are what matters -- after all, the entire point of this (and most if not all) article(s) here are for the benefit of people considering a purchase.
  • Lolimaster - Thursday, September 17, 2009 - link

    And then they talk about anand not being biased...

    PII 955 $189
    PII 945 $169

    Hey even in my country (Peru) wich has a lot of tech taxes the 955 costs less than AMD "official" price.
  • silverblue - Thursday, September 17, 2009 - link

    Depending on which sites someone uses, they may not offer a better deal than another site they may be unaware of, likewise they could get a CPU for cheaper than you or I can. It's strictly relative in the end and doubtful to be the result of some sort of bias. Anand seems very impressed with the Propus and so too will be a lot of other people.

    Looking more closely at it, I believe it should perform closely with a similarly clocked (or overclocked?) Phenom I due to the lack of an L3 cache, though I couldn't say which would perform faster. Incidentally, I've been checking dabs.com and their most popular CPU is the Phenom X4 9650 at about £80; this is only 2.3GHz and is more expensive than the newly-launched Athlon II X4 620 by about £5, so it stands to reason that people will go for this one instead soon enough.

    What I would really like to see, however, is a direct comparison between the 620 and the PII X4 905e (2.5GHz, 6MB L3) - it should be relatively simple to overclock one/downclock the other and see how much the lack of an L3 cache hurts the Athlon in a head-to-head, why the Athlon II X4's TDP is 95W whereas the 905e (with the 6MB L3 cache, remember) has a TDP of 65W, and whether it's worth the extra £50 for the 905e.

    I did look at other CPU prices on this site as well as on a couple of others. The PII X4 955 and the i5-750 are roughly the same at about £150-£155 (though dabs.com has the 955 as £165 for some daft reason), whereas the PII X4 965 is closer to £180. It's very important to look around for the best price, however I'm starting to doubt the reason for the 965's existence now, at least until they cut its price.
  • mapesdhs - Thursday, September 17, 2009 - link


    Indeed, shopping around is essential, and watchout for shipping
    charges aswell.

    After deciding what to buy, I Google for the part numbers or names
    of the items I want, gather a selection of sites, see what can be
    saved re shipping costs by buying multiple items from one seller.
    I've ended up using a variety of companies over the years, sometimes
    well known (Dabs, Scan, Microdirect), sometimes more obscure (Komplett,
    C&C Central, Lambdatek, Overclockers, Tekheads). And doing it this
    way means one has a better chance of locating special offers. I also
    check on eBay for BIN offers from reputable sellers, which in one
    case resulted in the best price for a PSU I wanted. The company names
    above are in the UK, but the same applies anywhere.

    Ian.

  • East17 - Thursday, September 17, 2009 - link

    1. A benchmark should test the architecture clock for clock. Dynamic overclocking just messes things up even if it's on by default.

    Sure Turbo is a nice addition and sure it should have been available ever since dual core CPUs appeared just to prove customers that you're doing something about single thread performance too.

    Turbo is a feature, not an architectural advantage. It's like saying a mainboard with Quad GigabutLAN on P45 is more performant than a Dual LAN on X58 just because it has this feature of having four LAN ports. Sure, if you do a LAN performance test you could find many ways in which you could show that the mainboard with Quad LAN is faster but surely you couldn't argue that P45 is superior to X58.

    Anyway ... the testing shoul have been done with Turbo on and Turbo off to clearly show thr difference .

    2.Deneb is 252% of Propus (300mln -> 758 transistors) but only offers around 10% performance improvement... Could you justify making the die size so much bigger just for 10%? Besides, if we talk about profit, if 252% = 225$ (Phenom 3GHz) and 100% = 99$ (Athlon II X4) ... Propus seems more profitable .
  • carniver - Wednesday, September 16, 2009 - link

    Athlon: Emergency Edition
  • silverblue - Wednesday, September 16, 2009 - link

    Sure, AMD has been forced to realise that competing at the higher end is a worthless exercise thanks to Lynnfield, however they've had Propus on their roadmap for a year now so I'd hardly term it "Emergency Edition". It's not a clock-speed bump like a certain Emergency Edition we all remember (how much did the various P4 EEs cost again?).

    With AMD providing a more complete platform, and especially one that allows for easy upgrades, the new Athlon II X4s could work for them in that it'll be cheap to upgrade from the old Athlon X2s as well as relatively inexpensive to build a new setup with CPU/motherboard/RAM (remember the onboard Radeon still runs rings around any Intel IGP). I doubt that i5/i7 sales make up most of their current order book (hell, they're still advertising Core2 on television and in IT publications here in the UK) so, like with graphics cards, the majority of the money stands to be made from mainstream and value products. As someone said before, this could be great for OEMs to improve their margins, as well as AMD if it means more bundles are sold. What's more, as things get more multi-threaded, dual cores fall more and more below the curve.

    I'm interested in seeing what Intel does, and even though their market share has grown again this quarter, I don't believe they can just do nothing about the Athlon II X4. Generally, below $150 or so, AMD does offer better performance for the price, even if, per clock, an Intel chip does more (and it's not always down to Intel-specific optimisations either) and earns its maker more money.
  • silverblue - Wednesday, September 16, 2009 - link

    Slight clarification on the P4 EE; it wasn't just a clock speed bump (3.2GHz -> 3.46GHz -> 3.73GHz) as, for example, the FSB was increased with the 3.46GHz model to 1066MT/s, then the core changed from Gallatin to Prescott (inc. extra cache) for the 3.73GHz model.

    Don't want anyone to think I just picked on the P4 EE just because it was fashionable or anything :)
  • Ratman6161 - Wednesday, September 16, 2009 - link

    Microcenter is advertising a special on the i5 750 at $159. But even at the full $206 suggested retail, if you are really doing something that benefits from 4 cores, it doesn't really make sense to me to go with anything less. There are even 1166 motherboards for just over $100 so the $99 quad just isn't saving enough to make it worthwhile.
  • gstrickler - Wednesday, September 16, 2009 - link

    On page 6 of the review the Excel Monte Carlo Simulation test shows the Q8200 (2.33GHz) being faster than the Q8400 (2.66Ghz)

    Same thing happens again on page 7 on the WinRAR Archive Creation test.

    You didn't mention it in the text and it's very unlike you to overlook or fail to comment on such an odd result.
  • flipmode - Wednesday, September 16, 2009 - link

    you said
    Overclocking suffers a bit as the chips capable of the highest clocks are destined to be Phenom IIs.

    And I think this does not make any sense. These are probably manufactured on their own wafers, and they are not Deneb dice, but completely different dice, so regardless of the clockspeed that any individual die is capable of, it cannot become a Phenom 2.
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    You're quite right :) I clarified in the article. That would apply to rebadged Denebs, but as to why the Propus core doesn't clock so well it's not totally clear to me at this point.

    It could be a design decision or just the impact of manufacturing a brand new die or just a bad sample.

    Note that I could get the chip a lot higher but it wasn't stable by my definition :)

    Take care,
    Anand
  • gstrickler - Wednesday, September 16, 2009 - link

    It's not overclocking. These chips are rated to run at "turbo" speed indefinitely. Each of the 4 cores is capable of running at that speed. However, for power and thermal management of all 4 cores, the on chip power management "underclocks" the cores dynamically depending upon load to keep the combination of all active cores from exceeding thermal and power limits.

    Intel is being very honest in listing the speed as the slowest speed at which all 4 cores can run concurrently, even though they're all rated to run 1.5x bus multiplier steps faster. The difference between the slowest "underclock" and "turbo" mode on the upcoming Core i7 mobile chips is reported to be much more than 1.5x (~6x). Again, that's for power (mostly) and thermal management, allowing it to operate as a fast single/dual-core or slower tri/quad core depending upon current usage.

    The reality is that most software can't currently take advantage of more than 1 or 2 cores, and for those applications, you're going to get the benefits of turbo mode. When you're running tasks that can benefit from 3 or 4 cores, the cores reduce their clock to limit the power and heat, giving lower performance on each thread, but higher total throughput.

    The user gets the benefits of a fast dual core and a slower quad core, depending upon current load. Because all of that power management is handled by the PM on the CPU, it's transparent to the OS.
  • The0ne - Wednesday, September 16, 2009 - link

    I see many angry users concern about the term overclocking. It strikes me hilarious that people really claim to know what Intel is and/or have actually done with their chip.

    I too agree it is not a fair comparison because the Turbo mode does increases the frequency. Granted that this is a feature of the chip itself but that is a feature that is not present on the AMD and hence isn't a apples to apples comparison. If you're going to represent Intel with a 600MHGhz boost why not compare it to a regular clocked AMD and an AMD chip that is clocked around the same turbo-boosted speed. That would be a fair comparison and would give anyone a clearer picture.

    The reason why the arguments by most of the users here are hilarious to me is that I don't think they have a clue to how the CPU's were tested. We have had CPU's that were generally locked to keep the speed/price ration. This "feature" was not presented as a feature to the consumer but once the market found out the chip could actually provide more then they in a sense enabled it. So the difference here now is is that Intel is making this available, unlocked and calling it a feature and suddenly cries of joy and stupidity are abound. Yes, yes, the technology isn't entirely the same but the end result is.

    Do you seriously think Intel did not overstress test their chips and fool-heartily tested only at stock. Do you think they didn't test them in turbo mode and still continue until the chip died? I can't recount the 30 years on my engineering career not having to test products in the extreme range of just about anything (temp, vibration, shock, emc, drop, shipping, customer use, etc.). If you are an engineer and don't do this you shouldn't be selling your shtty product, end of conversation. Feel free to disagree and be laugh at by the industry.

    Having said of of this the simple matter is that if Intel's feature is boosting the clock by 600, then why not compare an AMD chip with similarly clocked speed? And further more, why do you even rant about having to do this in the first place? Getting caught up in this "overclocking" terminology and not looking beyond what stands to reason is ignorant. At the very least I would hope to see the lower price Intel I5/I7 compare with the competition in default and turbo modes. Why wouldn't YOU?

    Flame away,
  • Ratman6161 - Wednesday, September 16, 2009 - link

    Weather its overclocking or not, leaving the turbo feature on the i5 makes it a good comparison for those people who want to just plug in the chip and run it within its specs. You are comparing "out of the box" speed of one versus the other. It doesn't make sense for every one, but I think the article gave enough information for the knowledgeable reader to make their own decisions.
  • pervisanathema - Wednesday, September 16, 2009 - link

    How can anyone advocate intentionally crippling the superior product in order to make the comparison more favorable for the competition? That's like a car magazine intetionally removing the turbo from a factory turbo charged vehicle so "that would be a fair comparison and would give anyone a clearer picture."
  • The0ne - Wednesday, September 16, 2009 - link

    Whether you like to believe it or not, crippling or, in a nicer way, not release some of the features is pretty common in both software and hardware environments. Whether the decision was due to money, managerial ignorance, or a time-line it is still a feature that is not in the product that could be there.

    If you consider putting all features, which is impossible btw, then you run into issues where the consumer or market doesn't even need them. We have USBs, wireless, biometric security available in one of our product but 95% of the market could care less. They are still on serial lines and are uneducated on newer technologies. Most city halls are like this that is why you see plain old switching voting systems still in place and the occasional typewriter.
  • The0ne - Wednesday, September 16, 2009 - link

    You mean like the I5?
  • andrenb91 - Wednesday, September 16, 2009 - link

    it's just a cpu...I'll buy it when I need it, my amd athlon 4850e and intel pentium dual-core e5200 still good enough
  • MrCommunistGen - Wednesday, September 16, 2009 - link

    Anyone know where you can get these? If not yet available, then when? I have a friend who needs a new computer, and the AII X4 620 would be a pretty good fit for him. I was somewhat grudgingly recommending the Ph II X2 550, but the X4 620 seems to offer much more balanced performance.
  • pervisanathema - Wednesday, September 16, 2009 - link

    These results are meaningless. The intel CPUs should be forced to run 24x7 at their highest possible turbo speed. To do otherwise let's AMD lose by a smaller margin.

    i'm going to stay here saying the same until hell freezes.
    i'm not going to accept under clocked results presented as if they were stock results.
    this is a casus belli.
    i mean it.
  • fitten - Wednesday, September 16, 2009 - link

    Heh... I see what you did there ;) I'm with you!
  • johnsonx - Wednesday, September 16, 2009 - link

    lol
  • ClownPuncher - Wednesday, September 16, 2009 - link

    I see what you did there! Nice job
  • Exar3342 - Wednesday, September 16, 2009 - link

    For $50 more, you get a much more energy efficient and faster processor. I would only recommend this quad to those with a MB that supports it, they need an upgrade, and they don't have much cash.

    AMD can't be making much money off these processors...
  • bji - Wednesday, September 16, 2009 - link

    "For $50 more" means "for 50% more" in this case. It's kind of ridiculous to talk about the difference of $50 in this context as if it's trivial. Those costs are not remotely comparable, so the implication that you might as well just spend $50 more to get the faster/more efficient processor is not justified by your statements.

    I personally always do target $150 for a processor price when building a new system - this is a habit I got into with my very first build, using a Pentium 100 which at the time cost $150. This means that a $200 processor which has better performance is not even in the realm of consideration for me, even though it is "only $50 more", which in this case, is an even smaller percentage increase than in the case you are talking about.

    Just in case I haven't made my point abundantly clear: you can't recommend spending $50 more for someone who has budgeted $100 for the processor. You have to compare similarly priced CPUs. I would like to see more comparisons with the Phenom II 550 BE for that reason.
  • Exar3342 - Wednesday, September 16, 2009 - link

    $50 is pretty trvial in the overall cost of the computer. Even a relatively cheap system with a O/S and a acceptable graphics card would be $400-500. This difference is only ~%10 of the overall cost, and yields a performance increase of 20-40% and is more efficient.

    You are thinking small, think bigger.

    With the Athlon X4 at $100 and the i5/i7 at $150-250, there is really no reason for anyone to buy a brand-new PhII system at all. If you want cheap, get the Athlon X$; if you want fast, get the i5/i7.
  • Patrick Wolf - Wednesday, September 16, 2009 - link

    "With the Athlon X4 at $100 and the i5/i7 at $150-250, there is really no reason for anyone to buy a brand-new PhII system at all."

    You are thinking big, think smarter. And the i5 is $200, not $150.

    The price per performance scales linearly. An Athlon X4 w/ mobo and DDR2 < PhenomII X4 w/ mobo and DDR2 < i5/i7 w/ mobo and DDR3.

    You get what you pay for.

    And less we forget, DDR3 isn't exactly cheap yet. And the new 1156 boards are starting at ~$100.
  • lopri - Wednesday, September 16, 2009 - link

    To all those who worry about AMD's finance: Why do you care?

    It has been a great mystery to me throughout many years. Sure I understand the need for competition which benefits everyone in a free market, but there are other things that can ensure fair competition. Worrying about a corporation's profit margin is not the first on the list. I'd leave that to the management and shareholders.
  • Smidge - Thursday, September 17, 2009 - link

    Competition is important, both for driving prices down, and driving development. As I recall, Intel started their whole tick-tock releases of architectures in response to starting to lose the lead in the processor market. As well as that they were reportedly sitting on a bunch of tech advances until they'd sold enough of their previous gen stuff. The cost savings are especially obvious when you think of how prices tend get slashed whenever the competition releases a new product, especially if it outperforms products in a higher price bracket.

    Now remember that AMD are the only real competition in both the CPU (x86_64) and graphics markets for Intel and Nvidia respectively. Both markets have an extremely high barrier to entry. Well, with the requirement of probably billions in startup investment and decades of processor research, it's more like a nigh-impenetrable barrier to entry. So if AMD were to go under, Intel and Nvidia would both have monopolies in their respective markets for a long time to come.

    I think it's perfectly fair to worry about AMD's finances given how much it would suck for us consumers if they were to go under (as they were close to doing before the globalfoundries spinoff). As I would worry if Intel or Nvidia were struggling. Though Intel seem to be sitting on a boatload of cash so that's not much of an issue.
  • BSMonitor - Wednesday, September 16, 2009 - link

    Just noticed on the benchies that the Athlon II at 2.8GHz was pretty close to the Phenom II at 3.2GHz... If you compared the two at the same clock, what does the 6MB L3 get one across the board?? In a lot of cases, seemed to be very little. Just one of those charts with percentages side by side would be cool!
  • strikeback03 - Wednesday, September 16, 2009 - link

    That chart is in the middle of the first page - a breakdown of the SysMark results at 2.8GHz.
  • BSMonitor - Wednesday, September 16, 2009 - link

    Right, that's just SysMark.. Sometimes he has a chart from top to bottom with all the benchmarks..
  • fitten - Wednesday, September 16, 2009 - link

    In the Cinebench R10 section:

    "the Q8200 is the slowest chip here." when it is clearly in the middle of the pack. I think it should be "the Q8200 is the slowest quad core chip here."
  • fitten - Wednesday, September 16, 2009 - link

    My bad... that is talking about the single-threaded performance, not the multithreaded performance.
  • DrMrLordX - Wednesday, September 16, 2009 - link

    I'm curious about the NB overclocking on these chips. Few have been able to achieve stable NB speeds over about 2.6 ghz on Phenom IIs. How far could you push the NB on your 630? Or 620, for that matter?
  • AznBoi36 - Wednesday, September 16, 2009 - link

    Typo on page 2.

    "Any strenuous video encoding however will seriously favor the Athlon II X4. Here we find the $99 620 tying the Core 2 Quad Q8200, and the 620 outperforming it - all at a lower price."

    Should be 630.
  • zivnix - Wednesday, September 16, 2009 - link

    We look at the TOTAL system power consumption.
    We look at the TOTAL system performance.

    Why do we compare prices of single components?

    If you consider TOTAL system cost, we don't look at 60% price difference. It falls to, what, 10%?

    And then even CPUs that cost more make sense.
  • flipmode - Thursday, September 17, 2009 - link

    Does not make any sense in my opinion. I already have a case, a PSU, a DVDRW, several hard drives.... I'm not going to be replacing those for no reason. If you want complete systems compared, go look at system builder prices.
  • mapesdhs - Wednesday, September 16, 2009 - link


    Anand, could you include a 3GHz AM2 Athlon64 X2 6000+ into the mix
    when doing these reviews please? It would be incredibly useful to
    know how the Athlon IIs and Phenoms compare to the dual-core
    Athlon64s (no need to compare to anything other than the 6000+). My
    older Asrock system is a 6000+ and the mbd can take the Phenom2 X4
    3.2GHz - but is it worth it? I don't know. Reviews keeping leaving
    out the Athlon64 X2, or if they do then it's some pointless low-end
    such as a 5600+.

    I expect Asrock will add BIOS support for these Athlon II X4s aswell,
    so again some comparison numbers would be good to know.

    And given earlier articles here and elsewhere, these new CPUs could
    also be a very handy upgrade for those trying to get the most out
    of AGP systems.

    Ian.

    PS. To everyone else: don't respond to the trolls. They merely seek
    attention. Replying just fans the flames and is exactly what they
    want. They know full well the points made back at them are correct,
    but that's not why they're posting. Best thing to do is ignore them.

  • Natfly - Wednesday, September 16, 2009 - link

    http://www.anandtech.com/bench/">http://www.anandtech.com/bench/
    The scores for the Athlon II X4s are up there.
  • subbotniki - Wednesday, September 16, 2009 - link

    Agree about including the X2 6000+ (great article though Anand).

    About trolls, AMD and the dragon: Myself is an AMD fan by consumer politital bias. I will always buy and use AMD. Therefore I love to read about a good product they have released. Beeing objective and fair, there are no ways today for an AMD chip to beat the Intel dito.
    Hmmm..here everyone knock their heads into the wall - including me: What are we going to compare? Which chip, what cost (incl power eg), plattform etc. Even I hurt my head quite a bit. I've been into hardware for 17 years now (RIP Cyrix!).

    The only real straightforward answer I give folk is: what (purpose) are you going to use the chip (computer) to/for? Next question is how big wallet you have and how much you are ready to spend. If money isnt a matter, then we don't need to speculate about whether superman or batman are the best; just buy the most expensive you can get hold of (and if that's not 'nuf - buy some more! In fact it isn't hard to build a system for $200 000. Just a money issue.. ).

    I guess most people doesn't have that kind of resources, and the one who has certanly doesn't write comments here. We DO care about prices
    and want to have as much as we can have for as little money as possible. Therefore I am convinced that when we messaure chips (consumer stuff) we have to look at the same price. It's not hard to imagine that a $100 000 car is better and fancier then a $10 000. It's also not to hard to understand that most people are going for the $10 000 car.

    But if the difference isn't a factor 10 but more like...hmm x 1.2?
    Back to basic: how big is your wallet and what are thoose xtra quids leave you with? Purpose again! I KNOW that most people can't tell the difference sitting in front of a 1.7 Ghz Sempron socket A and a core 2 duo 2.2 Ghz. Ever. I'm not talking to you computer freaks, I am talking ordinary stupid user here.

    And what I really miss here Anand(You AMD-freak! :-)): Where are the test under Linux enviroment? Take thoose champ for other purposes and you'll come up with some different results..whereas Intel arent fed by MS-platform. Cryptography, MD5 checksum, fileserver etc are test I really miss. I do a lot of video encoding (always in Linux) and would love to see charts from a Linux platform (Yea yea, Phoronix is da shit, but I simply love Anandtech).

    I'd better stop now - I know they throw thing @ AMD fans like me..

    Peace!
  • yacoub - Wednesday, September 16, 2009 - link

    If they launch one of these at a 45w TDP, it could be great for a small form factor system with a uATX mobo. HTPC...
  • arjunp2085 - Wednesday, September 16, 2009 - link

    Hi ,

    Was searching for Phenom X4 9850 Does not seem to be in the list .. Its not that old In my region where i live that's almost the price range the Phenoms sell for,,,

    If possible Please Try adding those charts Does Athlon X4 beat Phenom X4???
  • blyndy - Wednesday, September 16, 2009 - link

    That's a bloody nice deal. So it's only about 70% as powerful as the top of the line processors. It has more than enough processing power for 90% of computer users, and it can handle all the latest games.

    P.S. "Overclocking suffers a bit as the chips capable of the highest clocks are destined to be Phenom IIs" I can understand that they'll turn a few crippled Phx4s into A2x4s, but why would overclocking suffer on a deneb A2x4 just because some cache has been disabled? Can you please clarify this?
  • PrinceGaz - Wednesday, September 16, 2009 - link

    Doesn't the part "the chips capable of the highest clocks are destined to be Phenom IIs" answer your PS? They test roughly how fast each chip can run, and bin them accordingly. The faster ones end up as Phenom II, the not so fast ones become Athlon II.
  • blyndy - Friday, September 18, 2009 - link

    I don't know if they intentionally cripple fully functioning denebs. I imagined that amd would rather sell a Phx4 for ~$200 than an A2x4 for ~$100, but I might have read somewhere that both intel and amd do intentionally cripple fully functioning chips.
  • strikeback03 - Wednesday, September 16, 2009 - link

    That statement makes sense for harvested Denebs, but doesn't Propus not have any L3 at all, so it can't be turned into a PhenomII?
  • blyndy - Friday, September 18, 2009 - link

    Yes that's why I thought the quote didn't make sense -- If propus is exclusively A2x2 then the highest clocking propus' will still be A2x2's, therefore A2x2 overclocking wouldn't suffer.
  • MrPIppy - Wednesday, September 16, 2009 - link

    How is the motherboard compatibility situation looking? Will boards need a BIOS update, or just work out of the box right now?

    Also, does it support AMD virtualization instructions?

    Last, any idea about compatibility with ECC RAM? The BIOS often plays a role in this too, but just wanted to make sure AMD didn't remove ECC support from the IMC or something similar
  • TA152H - Wednesday, September 16, 2009 - link

    This is a GREAT product from AMD.

    Make it smaller, make it cheaper, and put it with a platform much better than one Intel can have.

    AMD can not make a processor worth a damn. Let's face it, they suck. If they try to compete head on with Intel, they lose, period.

    By making a quad core that is more than fast enough for most people, while at the same time reducing the size so they can make it cheaper, they created something in a segment where Intel just isn't.

    Couple this with a 790GX, or the 785, and you've got a great platform for a lot of people. If you need the best, or near it, the Bloomfield can't be touched. Why even try? Most people don't need it.

    Finally, AMD seems to get it.
  • khaakon - Thursday, September 17, 2009 - link

    I kinda agree with you here. But I cannot understand your need to paint the world in only black and white.
  • khaakon - Thursday, September 17, 2009 - link

    * edit;

    "your need" meaning TA152H
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    "Finally, AMD seems to get it."
    Yes, but will they make money out of it? Only time will tell.
  • Chlorus - Wednesday, September 16, 2009 - link

    Where's that dipshit snakeoil to complain about nonexistent bias?
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    He's in disguise under a new screen name. See if you can spot him in the comments so far.
  • Chlorus - Wednesday, September 16, 2009 - link

    He posted right after me...I'm kinda wondering if we should setup a drinking game or something each time he posts.
  • strikeback03 - Wednesday, September 16, 2009 - link

    Good idea, and everyone drinks when you spot a new alias.
  • deputc26 - Wednesday, September 16, 2009 - link

    I would absolutely love to see a die size comparison. That will give real information on how this chip can make AMD $$.
  • Lokinhow - Wednesday, September 16, 2009 - link

    there is a die size comparsion on the first page..

    P2 X4 = 258 mm2
    A2 X4 = 169 mm2
    A2 X2 = 117 mm2
    C2Q 8xxx = 164 mm2
  • deputc26 - Wednesday, September 16, 2009 - link

    Apparently I'm a moron. Don't know how I missed that unless it was added after initial publication as I read it right after it came out.
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    Anand, how did you know whether or not your Athlon II X4s are harvested Denebs?

    Also, any idea why the Athon II X4s debut at such a high TDP with no L3? I'd think they'd be lower at 65w or at least 80w.

    AMD, $100 quad cores is nice... BUT... where's the 45w quads???
  • Lokinhow - Wednesday, September 16, 2009 - link

    I may be wrong, but..

    I think the high TDP is because we have both Propus and Deneb-Harvested Cores

    The Athlon II based on Deneb probably have this 95W TDP, while the ones based on propus core have lower TDP.

    I think they have only a few propus core to sell, so that's why they are selling harvesteds Denebs. When there wil have only Athlons II X4 propous based they'll rate it with a lower TDP.

    Makes sense?
  • Spoelie - Wednesday, September 16, 2009 - link

    Yes and no

    It's a bit strange that the deneb die (630) consumes less power than the propus die (620) in idle if that theory was valid. During load the difference is just as much as the difference in clock speed would indicate. So if the 630 is indeed a deneb (care to rip the IHS of? ;)) then this means that propus is not by definition less power hungry than deneb.
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    AMD specifically told us that our 620s are Propus samples, but that others may be Denebs. I'm not sure how to tell if you have a harvested Deneb just yet.

    Give AMD some time, I'm sure we'll see them down below 95W as the process matures for these dice.

    Take care,
    Anand
  • TA152H - Wednesday, September 16, 2009 - link

    Anand,

    A couple of things. You mentioned the 'small' L2 cache being a problem on one of your benchmarks, but, it's actually twice as large as the one on the Phenom. I'm not sure if this was meant to be a comparison only with the Penryns, but it's a bit confusing.

    Also, going back to the L2 cache, how can these possibly be harvested from Phenoms with a bad L3 cache. That would imply the Phenoms are built with 512K L2 cache, with half of it disabled. I really doubt this is the case. You CAN remove the L3, but how do they then double the L2 cache? This seems strange to me.

    Based on the relatively poor overclocking potential of this chip, would you attribute that to the L2 cache? Does the L2 cache run with the same number of wait states as that of the Phenom? If so, that could prove to be the main reason for the lower overclocking potential. Any ideas on this?

    Also, don't you think it's worth mentioning AMD's greatly superior IGPs, considering this product could easily find it's way in this platform rather often. The processor by itself does make sense, but, even if it didn't, the superior IGP platform still can make AMD processors somewhat attractive.
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    The L2 cache in the Athlon II X4s are the same as the Phenom II X4, 512KB per core, 2MB total. It is only on the Athlon II X2s that the cache was doubled to 1MB per core, 2MB total.
  • TA152H - Wednesday, September 16, 2009 - link

    Oh, OK. I guess I should actually read the charts.

    Thanks.

    Hmmmmm, I wonder why they overclock so poorly. It doesn't make much sense. You'd think it would use less power, and generate less heat. Strange.
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    Two options: 1) New die, takes time to get the mix perfect for better yields/higher clock speeds, or 2) the chip isn't using super high frequency/high leakage transistors to maximize performance. It could be designed to hit lower frequencies.

    Take care,
    Anand
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    The skeptic in me says bad design problems, like the first Phenoms. That could also account for the multiple delays and pushbacks on the Athlon X4s. But in all seriousness, I have no idea.
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    Found this on another site. Perhaps you can check and confirm:

    "...the imprint "AADAC” identifies the CPU as a Propus."
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    My 620 is an AADAC while my 630 is an AACYC. I will ask AMD to confirm :)

    Take care,
    Anand
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    Spoke with AMD this morning. The OPN does not indicate whether or not the chip has a disabled L3. It's just luck of the draw, there's no way to tell by looking at the chip itself.

    Take care,
    Anand
  • Doormat - Wednesday, September 16, 2009 - link

    Please upgrade to a more modern x264 benchmark. I'd recommend a recent handbrake snapshot (http://handbrake.fr/snapshot.php)">http://handbrake.fr/snapshot.php). The nehalem optimizations should boost performance dramatically and are a better representation of what people would get with current encoders.
  • Anand Lal Shimpi - Wednesday, September 16, 2009 - link

    I agree completely, the question is more one of when we make the transition. There's a lot of historical data we need to compare to. You'll see a slow transition to new tests especially with the final version of Windows 7.

    Take care,
    Anand
  • Lunyone - Wednesday, September 16, 2009 - link

    Finally we have quad cores at about $100!! Hope that this will spur on better pricing on all fronts! Go AMD, we need the competition to drive down better pricing. Now only if you would drop down the PhII x4 955/965 pricing to compete better with the i5 750!!
  • thezorro - Wednesday, September 16, 2009 - link

    remember that the lynnfield core i7 750 has turbo overclocking enabled so is overclocked to 3.2 ghz, which is 600 mhz overclocking, while amd processors are running at stock speed
    Is unfair to present the results this way, turbo is overclocking.
  • Wivvix - Thursday, September 17, 2009 - link

    Sorry but your logic fails. It is a feature of the i7 750, that it runs at slower speeds when idle, as to when the processor is under load.

    In the same way, fan speed RPM is automatically regulated depending on whether the process is idle or under load.

    The maximum clock speed of an i7 750 is 3.2gh. The user doesn't have to change or do anything. This is not overclocking in the sense you purport, and is not unfair. That is the maximum clock speed of the processor out of the box. End of story.
  • kagenokurei - Wednesday, September 16, 2009 - link

    im sorry, but i have to defend thezorro at least this time..

    overclocking's basic principle is to increase CPU speeds..but if u guys look into it, what u increase is the core clock(GHz,MHz,etc) beyond the factory released specs..voltage increases and other tweaks are only done to make the overclock stable..

    now,in this light,i also find it quite unfair to compare the single threaded performance of the i7s with the PII X4s..simply because the i7s can run its single core at a higher clock when the others are idle..therefore,overclocking at least THAT core..w/c of course, the PII X4s are unable to do..

    i agree that the Turbo mode is a good feature, and AMD should have something like it on their CPUs..until then, the Turbo mode is kind of an unfair advantage in benchmarking single threaded performance..
  • jonup - Wednesday, September 16, 2009 - link

    Well the PII 965 is an overclocked to 3.4GHz, which is 600MHz overclocking of 920. It's unfair to present the results this way. The 965BE is so obviously OCed because it uses tons more power than the 920.
  • fitten - Wednesday, September 16, 2009 - link

    Bzzzt... wrong... when will you fanbois give this up? It's the normal and default behavior of the thing. You can do the academic discussion all you want but normal operation of the chips has Turbo enabled. To test any other way is to test a non-standard, non-default operating mode. You might as well test with the caches disabled or one or more cores disabled since those are all non-standard, non-default operating modes as well.
  • maxxcool - Wednesday, September 16, 2009 - link

    ++ ding!
  • Archangel59 - Wednesday, September 16, 2009 - link

    I can see both sides of this where Turbo could be considered altering the "normal" operation and also where turbo could be considered just the way it works.

    Doesnt bother me either way that this is put in the article, still good news from AMD.

    However, since the turbo mode could be considered a supplied feature of the chip... couldnt the use of AMD Overdrive be considered a supplied feature of the newer chips? Even if it couldnt, I'd still love to see an article comparing the turbo mode to a well set up AMD Overdrive profile.
  • Lunyone - Wednesday, September 16, 2009 - link

    So what your saying is that the test are fair because AMD doesn't have this "turbo" mode, so all of the benchmarks are created equal?? So if the benchmarks had set up the 955/965 to 3.2 gHz (with unlocked multiplier) than you would have a favorable review??
  • Lunyone - Wednesday, September 16, 2009 - link

    Crap! I meant to say "aren't fair" in a couple of lines in my response :(
  • SlyNine - Wednesday, September 16, 2009 - link

    Turbo mode is not overclocking. Overclocking is clocking OVER factor speeds.

    If the CPU can increase the speed as a factory set speed. THAT IS NOT OVERCLOCKING, so get over it.

    On the other hand. Go AMD. I'm still awaiting the big Core I7 killer. comon AMD you can do it.
  • smn198 - Wednesday, September 16, 2009 - link

    [quote][It] is unfair to present the results this way, turbo is overclocking[/quote]

    I think we need to review our definition of overclocking. For years both Intel and AMD, have varied processors' speed and voltage based on temperature and workload. Intel has now changed how they are doing this. Using two cores your processor runs at one speed and using all four, it runs at another. As this is how Intel is selling the CPU I'd argue that this isn't overclocking as by definition, this is running at a frequency beyond spec. Therefore it is a perfectly fair way to present the results.
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    "Therefore it is a perfectly fair way to present the results."
    - Yes, but it's also an incomplete way of viewing Turbo CPUs, as demonstrated in Gary's follow up benchmarks. Based on his benchmarks, the i5 loses its Turbo advantage if you start heavily multitasking the system. In heavy multitasking, the i5 performance approaches the performance of i5 with Turbo off, which was generally worse than the Phenom II 965, at least based on the benchmarks supplied by Gary. So it's definitely another variable to keep in mind when comparing CPUs.
  • jonup - Wednesday, September 16, 2009 - link

    Well Intel could not clock the i5 at higher stock because it would have outperformed the i7s. People should stop complaining about benches with turbo on. Fact of the matter is i5 outperforms the PII clock for clock and overclocks higher that the PII. the 965BE is clocked 700+MHz higher than the i5 and it consumes a lot more energy. The fact of the matter is that Intel will not clock the i5 at 3.2GHz and disable the turbo mode (because of the i7), but you can do it (safely).
  • Eeqmcsq - Wednesday, September 16, 2009 - link

    > People should stop complaining about benches with turbo on.
    Well, it depends on what the complaint is. For me personally, my complaint has been that the Turbo on numbers alone are not enough to describe the capabilities of a Turbo CPU. I wanted Turbo off numbers too, to get a baseline on what the CPU can do when Turbo can't help.

    Thankfully, Gary has done that and it shows that when you throw a lot of tasks at the i5, it falls noticeably behind the 965BE. And that's important to know. If I had to choose between the i5 and 965BE at home, my choice is the i5, because I do so little at home that pushes 4 cores, that the Turbo would be useful and helpful. At work, however, I'm pushing all of my available cores compiling large amounts of source code, while zipping or unzipping files, while my version control client app is busy doing whatever it's doing in the background eating up one of the cores at 100%, while I'm running a simulator app which requires another core. In this case, I MUST choose the 965BE over the i5, because the i5's Turbo can't help me, so it performs closer to Turbo off, making it worse than the 965BE.

    So for me, that's why I want Turbo off numbers. I'm after the complete picture.
  • jonup - Thursday, September 17, 2009 - link

    Well as erple2 said make your decision based on the multithreaded benchmarks. Also, even in multithreaded apps the turbo kicks in (marginally), therefore, disabling the turbo is not giving you an accurate picture.
    Anyways, it seems like in you work environment you will be better off with an i7, because Gary's number do show significant improvement with HT on in multithreaded test. Especially since you use it for work, the ROI will justify the additional $100 expense.
    My point however was that if you are given the opportunity (bios access) you will be better off OCing the i5 to 3.2GHz (the single core turbo on value). After all 965BE is an OCed PII; it is obvious from the power consumption numbers. And due to the reasons I mentioned you will not see a factory OCed i5 (3.2GHz stock clock). That does not mean that i5 cannot run 3.2GHz turbo off (safely), and out perform the PII. I would actually speculate that @ 3.2GHZ, i5 will be within the thermal envelop of 965BE.
  • Eeqmcsq - Thursday, September 17, 2009 - link

    > Anyways, it seems like in you work environment you will be better off with an i7
    You are right and I agree that the i7 with hyperthreading would be the BEST choice for my work loads. Convincing an IT guy that I NEED an i7 when all of our programmers can adequately get work done with our old Pentium D's, uh, that's another matter.

    > you will be better off OCing the i5 to 3.2GHz (the single core turbo on value)
    Yep, this is also an alternative for me, but it works around the issue on how to compare Turbo CPUs to non Turbo CPUs. Also, see my, uh, "essay" written above.
  • jonup - Thursday, September 17, 2009 - link

    "Convincing an IT guy that I NEED an i7 when all of our programmers can adequately get work done with our old Pentium D's, uh, that's another matter. "
    Well, in that case you might have an issue convincing them that you need a 965BE. Or a i5 for that matter. :)
  • Eeqmcsq - Thursday, September 17, 2009 - link

    > Well, in that case you might have an issue convincing them that you need a 965BE. Or a i5 for that matter. :)

    Haha, yeah, I know! I'm stuck with my Pentium D for the foreseeable future. :( Unless our IT guys will bite at a $100 Athlon X4... and a bagel or two. :P
  • maxxcool - Friday, September 18, 2009 - link

    heh, I gutted my workstation after hours and installed a Kuma 7750 onto a 8200 series asus board. i *had* a p4 3.0ht and it was killing me inside when doing log file grep'ing, compression and local database work.

    the best part was using the companies "reward dollars" to buy the mobo and cpu. :) now.... I magically can do *more* work and am getting more reward bucks.... :D
  • jonup - Thursday, September 17, 2009 - link

    Well maybe if you get the begels u can talk him into it.
  • erple2 - Thursday, September 17, 2009 - link

    Ditch Tortoise, and you'll notice a speed boost. Or at least get rid of the icon overlays. That's what's probably killing your system. "Show modified files" is what you want (or use the command line client).

    Or, get an SSD.

    BTW, the complete picture would actually be heavily threaded benchmarks, not arbitrarily disabling turbo mode. I'm not constantly compiling at work, however. There are lulls now and then. So I would rather have a heavily threaded workload benchmark than arbitrarily turning off Turbo mode.
  • Eeqmcsq - Thursday, September 17, 2009 - link

    >BTW, the complete picture would actually be heavily threaded benchmarks, not arbitrarily disabling turbo mode

    Yes and no. Heavily threaded benchmarks would definitely add to the picture. But I've been thinking about this a lot, and I STILL believe that the Turbo off measurement is very significant, even for single threaded benchmarks.

    Whenever reviewers run benchmarks, they were always run isolated, meaning with 0% background tasks while the benchmark is run. This gives them the most accurate measurement numbers of a benchmark, thus an accurate ranking of which CPUs are the "winners" and "losers". If the same benchmark was run with 100% background tasks, we obviously know that the measured numbers would be WORSE than the 0% background benchmark, but we have always ASSUMED that the RANKINGS won't change. That's because the CPUs ran at a constant clock speed, so we assumed that the performance degradation would be equal and linear for all CPUs as the background level scales from 0% to 100% CPU load.

    For example, looking at the Cinebench single threaded benchmark here in this Athlon II X4 article, we know that the Phenom 955BE (score: 3675) wins over the Core 2 Quad Q8400 (score: 3198). This benchmark was most likely run with 0% background tasks. If the same test was run with 100% background tasks, the scores would certainly be much worse, for example, maybe 955BE scores 2000 while the Q8400 scores 1700. Although the score degraded, the RANKING of these two CPUs should NOT change. In other words, we ASSUME that no matter what the background load is, every time you run the Cinebench single threaded benchmark, the 955BE will ALWAYS beat the Q8400.

    With the i5's Turbo mode, since the i5 runs at VARIABLE clock rates depending on the background load, the assumption of linear degradation is NO LONGER ACCURATE. As you increase the background CPU load from 0% to 100%, i5 benchmarks will degrade BOTH from having a loaded CPU AND from the i5 reducing the clock rate down towards the Turbo off rate. It is possible for the i5's degradation to be so sharp, it will rank BELOW the 965BE. As a result, the SAME BENCHMARK can produce TWO DIFFERENT WINNERS depending on the background load!

    For example, from Gary's MainConcept Reference benchmark, his benchmark says the i5 Turbo on finishes in 249 seconds, the 965BE in 269 seconds, and i5 Turbo off in 280 seconds. The i5 Turbo on ranks ABOVE the 965BE, presumably because this benchmark was run with 0% background load. If the same benchmark was run with 100% background load, we know we will get worse times, but we also know the i5 will degrade and perform close to its Turbo off ranking. And for this benchmark, the Turbo off rank is BELOW the 965BE. So the winner of MainConcept Reference is either the i5 at 0% background, OR the 965BE at 100% background. The SAME BENCHMARK can claim that BOTH the i5 AND the 965BE are winners.

    This is why we need BOTH Turbo on AND Turbo off numbers for every benchmark. If the 965BE ranks ABOVE the i5 Turbo on, the i5 can NEVER beat the 965BE, so the 965BE is the guaranteed winner across all background loads. If the 965BE ranks BELOW the i5 Turbo off, the i5 can NEVER perform worse than the 965BE, so the i5 is the guaranteed winner across all background loads. If the 965BE ranks BETWEEN i5 Turbo on and Turbo off, the winner is a toss up DEPENDING on background load.
  • thezorro - Wednesday, September 16, 2009 - link

    i'm going to stay here saying the same until hell freezes.
    i'm not going to accept overclocked results presented as if they were stock results.
    this is a casus belli.
    i mean it.
  • Venatici - Friday, September 18, 2009 - link

    "i'm going to stay here saying the same until hell freezes."

    Or at least until AMD copies Intel and includes Turbo on all of their chips, which is exactly what AMD intends to do.
  • maxxcool - Friday, September 18, 2009 - link

    I am still shocked that with all the cross licensing between the two that amd has not come up with a hyper-threading style feature. but yes amd will copy this in short order.

    dynamic coring is a great feature idea... and both companies should be using it.
  • hob196 - Thursday, September 17, 2009 - link

    Ok, I'll bite.
    I, like most computer consumers, just want to know what it will do out of the box without having to do potentially dangerous things in the BIOS.

    Maybe turbo is 'dynamic overclocking'. Perhaps we should only compare all chips running at the same voltage and speed?
    But in all honesty I just care what I get for my buck without having to muck about.

    Ever wonder why we didn't see people ranting in this way about efficiency figures when AMD introduced dynamic voltage variation?
  • Nfarce - Wednesday, September 16, 2009 - link

    "i'm going to stay here saying the same until hell freezes."

    Well you can stick your head in the sand, or like a stupid child plug your ears, shut your eyes, and scream "nah nah nah I can't hear you!" As has been mentioned countless times, Intel has had "turbo" modes on many CPUs over the years. Overclocking in a traditional sense is upping bus speed, changing multipliers, modifying voltage, and all of the above or a combination of the above.

    You AMD fanboys can go cry a river about Intel having that advantage. AMD chose not to have similar technology. That's AMD's fault, not Intel's, and apples are apples when not overclocking in a traditional sense, as in the case of i7's turbo boost. You whiners can cry about it until hell freezes over is more like it.
  • maxxcool - Wednesday, September 16, 2009 - link

    will you die you douche bag stir stick? Its a default feature. every consumer in the world that gets a Dell computer based on i5/i7 will get the same performance performance boost. if they get a i7, it will be even more.....

    bitch all you want. every consumer in the world that uses a i5 will say the same thing... Gosh it's sooo fast and pretty.... and oh yes. it uses less power, saves the wales, promotes the rain forest cures herpes, and gets me out of 105mph speeding tickets in a school zone.

    sarcasm aside. you will be banned again soon enough just like you got BANNED on techreport you moron.

    and and don't forget your the IDIOT that screamed in our forums that the i5 would not support VMWARE or XP-MODE... you tool-bag of suck.








  • tpi2009 - Wednesday, September 16, 2009 - link

    Oh look! Hell just froze!

    Now serious, you can say what you want, but by definition, overcocking is something that is done outside the factory specifications and normally voids the warranty. The manufacturer does not guarantee that the processor will work reliably (or at all) in the short or long term at a higher clock speed than the one they established as good for that processor.

    What the i5 and i7's do has been thoroughly tested and is in fact "by design", it's guaranteed to work when it works, because Intel has a monitoring chip alowwing the frequency to be officaly raised in certain circumstances.

    On the other way around, Intel already did this kind of frequency switching, but much less elaborate, with EIST and AMD too, with Cool,'n' Quiet.

    Just because AMD hasn't gotten there yet, you can't say thatover at Intel it isn't a feature, because that exactly what it is, a feature.
  • oldscotch - Wednesday, September 16, 2009 - link

    Remember how an 80286 would run at 6 Mhz? There was a "Turbo" button that would boost that thing all the way to ... 8mHz.

    ZOMG OVERCLOCK!
  • fitten - Wednesday, September 16, 2009 - link

    Your definition of "overclocked" is flawed. Back under the bridge with you!
  • philosofa - Wednesday, September 16, 2009 - link

    Can someone please ban this moron, and the others (who ironically I assume are on AMD's payroll?). Their lack of an understanding of the most basic logic is making me feel ill.

    Good Review though, just really want Bulldozer to come out!
  • james jwb - Thursday, September 17, 2009 - link

    why can't people like you understand what is meant? I can handle my own CPU, so for me, i want to know which one is better clock for clock, and then see what the average each overclock to, then i'll jump in and buy one.

    The way data is currently being presented here isn't right, we need both ways (stock results with turbo on, and clock for clock style stuff for us overclockers).

    Get it now? Probably not.
  • bupkus - Wednesday, September 16, 2009 - link

    "Their lack of an understanding of the most basic logic is making me feel ill."

    Take a Tums; we're not interested.
  • rennya - Wednesday, September 16, 2009 - link

    Overclocking is not the same as turbo mode.

    Overclocking voids your warranty, whether you use Intel or AMD CPUs.

    Turbo mode doesn't void warranty because it is a valid feature.
  • SlyNine - Wednesday, September 16, 2009 - link

    lol You will make your own little universe in your mind. okay lol.

    Overclocking. Lets see. Over, meaning above normal. Well since the post turbo mode clock is perfectly normal. Its not OVER clocking.
  • vol7ron - Wednesday, September 16, 2009 - link

    Just because the clock rate is variable, does not mean it's [over/under]clocking. It just means the speeds change. In fact, even when you are running at stock speeds, the clock rate will vary - that is why you see 3.541 3.500 3.489

    To complain it's not a fair comparison, he's kind of right, but not for the right reason. It's as if he's in AMDs defense, rather than their scrutiny. Instead of comparing peaches-to-peaches, now you're comparing nectarines-to-peaches and there's two ways to look as this glass: Intel is turbo-ing during fewer threads, or Intel is decelerating in multithreaded situations; both have the same effect. (They're either trying to give you more bang for the buck, or they put out a bad product that only works half the time)

    The end result, though, AMD doesn't do this on the fly. Sure you could test it at different clock speeds, but you could also overclock the i7 to something faster (and you might be back at square one).
  • Chlorus - Wednesday, September 16, 2009 - link

    "remember that the lynnfield core i7 750 has turbo overclocking enabled so is overclocked to 3.2 ghz, which is 600 mhz overclocking, while amd processors are running at stock speed
    Is unfair to present the results this way, turbo is overclocking. "

    Its not overclocking if ITS ON BY DEFAULT, you worthless troll.
  • hanhan1982923 - Wednesday, April 14, 2010 - link

    22222222222222
  • Thomas369 - Saturday, July 28, 2018 - link

    The Athlon architecture also used the EV6 bus licensed from DEC as its main system bus. Intel required licensing to use the GTL+ bus used by its Slot 1 Pentium II and later processors. By licensing the EV6 bus used by the Alpha line of processors from DEC, AMD was able to develop its own chipsets and motherboards, and avoid being dependent on licensing from its direct competitor. https://chasebanklogin.us/chase-credit-card-login/

Log in

Don't have an account? Sign up now