Back to Article

  • niva - Wednesday, March 19, 2014 - link

    These benchmarks are making me depressed for AMD CPUs. I guess it's time to switch to Intel after not having purchased an Intel chip since 1996. Reply
  • nathanddrews - Wednesday, March 19, 2014 - link

    Come on in, the water's fine. Reply
  • Malorcus - Wednesday, March 19, 2014 - link

    I hear you man, I did the same with my current Ive Bridge CPU. I am looking to build a media computer using an AMD APU though. They still have their niche, but it is not in high end computing/gaming. Reply
  • ddriver - Wednesday, March 19, 2014 - link

    You'd be surprised by the amount of needed and commercially viable tasks for which those poor CPUs are more than fast enough. It is a sad thing to see AMD struggling to compete with Intel's value products. Reply
  • Fallen Kell - Thursday, March 20, 2014 - link

    AMD has sadly not had a real competing product in the high end side for 6 or 7 years now. On the low end, AMD was competing, until Intel decided to compete in this market segment. The last two updates that Intel has made were more focused on the lower end than on the high end. This is finally cutting into the one thing keeping AMD alive. I really hope AMD does survive because those of us that are old enough to remember know that Intel hates the consumer and only really pushes technology when it is competing. We would have CPU's that are only soldered directly into motherboards with no ability to upgrade, completely locked down CPUs with no ability to overclock, locked in memory bus speeds that are tiered based on the CPU/motherboard that you purchased with higher performance memory compatibility costing you extra, etc., etc....

    But I really don't see a way that AMD can compete at this point. They are still hemorrhaging money (not nearly as bad as a year ago when they lost $1.2 BILLION, but even after restructuring to cut 31% of their operating costs, they still lost $162 million last year). While I understood the reason for acquiring ATI, I believe ATI is worse off due to that acquisition. ATI went from being a profitable company competing well in their market, to one that is losing money and is seemingly almost a generation behind Nvidia in their offerings (I say this based on the fact that the brand new released top of the line AMD graphics cards can barely beat the last generation of cards from Nvidia in performance and can not come anywhere near the Nvidia offerings in power/performance or heat/performance, and Nvidia is getting ready to release its true next generation of cards even while they simply released the last generation of cards at their full potential to beat AMD's current cards now that AMD finally had a competing product). The major losses that AMD as a whole has, is taking its toll on the R&D AMD can afford to do in terms of increasing the efficiency of their designed with respect to power and cooling requirements while still being able to push the performance of their cards.
  • Demiurge - Friday, March 21, 2014 - link

    Funny, I thought the same thing about Intel CPU's when GPU's started to encroach on the high performance features such as physics, ray-tracing, signal processing, and other high intensity applications. There's a bigger picture that I think a lot of people miss. The market has already shifted away from CPU's being the centerpiece of high performance applications. AMD has the right strategy with buying ATI and the paradigm of Heterogeneous Computing, but like Intel with the P4: it's too little too late. If they had the software, they might've been able to pull this off, but that is exactly what they are trying to do with Mantle. I think only unanimous adoption would've guaranteed a win. It was a big risk, and it would have been an amazing upstart, but I don't think it will pay off as much as they need it to. Just a modest opinion... and so a long rant ends... ;-) Reply
  • pandemonium - Thursday, March 20, 2014 - link

    AMD hasn't been competitive to Intel for the consumer since 1999 or so? They've always been cheaper, and always been far lower in gaming and general desktop usage results as well. You are very, very, very late to the party, lol. Reply
  • mr_tawan - Thursday, March 20, 2014 - link

    I believe it was Intel Core series (2006) that started to get ahead of AMD's CPU. Before that, AMD CPUs was both perform better and cheaper. Intel CPUs were power hungry and expensive while did not yield excellent performance.

    The raw ALU performance on the current line of AMD CPU is quite low because of the design decision to reduce the space occupied by the CPU while add even more GPU on the die, and then make them work together more nicely. It's the direction the AMD heads to.

    I believe that one day even the FX line would be APU just like those A-series.
  • Vinny DePaul - Thursday, March 20, 2014 - link

    I feel your pain. I am an AMD fan. I want to keep using AMD but the CPU is running too hot. It is heating up the room! I switch to Intel. It is just easier.... I hope the AMD's involvement in PS4 and Xbox One will shape the future in games and software. Reply
  • Xpl1c1t - Thursday, March 20, 2014 - link

    This is exactly what I did going from a Athlon XP Palomino, to Athlon 64 Venice, to a Core i5 Lynnfield.

    I'd consider purchasing an AMD processor again if the whole APU thing becomes quite competent and powerful at a smaller and more efficient node. I'd promptly and gladly buy an APU with the equivalent of 2 IVB cores and a 7870 onboard if it could me mounted on a Pico-ITX board...
  • fteoath64 - Friday, March 21, 2014 - link

    AMD/ATI needs to be very aggressive in lowering the power demands of their gpu. What NV has done in Maxwell should tell them a very important thing. Lowering the power demands means, one can cram more cores into the die, hence it will boost performance. There is so much one can cram into a specific node-technology but the power demands meaning heat dissipation is going to be a real issue that is hard to solve.
    With the latest R9 series running at such high heat and high power demands, it is going to impossible to cram even half that performance into an APU without resorting to water cooling so the challenge is huge for AMD to tackle. I hope they can make headway into power optimization so that we can get more serious APU chips with powerful gpus for once and help move the industry along.
  • Ammohunt - Thursday, March 20, 2014 - link

    cpu performance is one thing getting a stable glitch free windows system has always been the true challenge. I had never seen a Win7 BSOD until i ran an FX-6100 AMD Build which i promptly replaced with an intel rig. As a server it runs linux like a champ though. Reply
  • Demiurge - Friday, March 21, 2014 - link

    I have multiple laptops, all Intel based, at work that BSOD Win7. It isn't the CPU, it's the drivers. I know because mine doesn't crash anymore after updating the drivers. I'm pretty sure it isn't the CPU that was the weak point of failure in most problems because the CPU is one of the few things that gets tested and validated the most. Not saying it can't happen, but it is far more likely you or the device manufacturers screwed up. Trust me, I own a Creative X-Fi =)... sometimes I don't get sound on reboot. I know the hardware's good there too. Reply
  • MrBungle123 - Friday, March 21, 2014 - link

    I've noticed instability with AMD rigs before and discovered that the issue was that the stock voltage was too low. AMD is always trying to compete from a process node behind so I think they drive the Vcore as low as possible to try and bring down their TDP numbers and as a consequence bring the CPUs to the edge of instability. Reply
  • 0ldman79 - Saturday, March 29, 2014 - link


    I've found that most of my AMD chips can work fine with a little less voltage at stock speeds.

    Of course they don't stay that way, the tech in me overclocks them and gets Cool'n'Quiet working so they idle nice and cool then ramp up when needed.
  • Demiurge - Friday, March 21, 2014 - link

    Not sure why...I see something a little different...

    Tomb Raider is probably very heavily AMD CPU (just assuming relative to the other benchmark performance) optimized because the $200 APU + GTX 770 is actually outperforming (ever so slightly) the $1000 and $320 I7... In the F1 game the exact opposite is happening ~84fps vs ~129fps... from what I see in the other games, it looks like about a -15% typical difference toward the APU just by looking at a high level. That's not bad considering the price/performance ratio.

    The numbers (to me) aren't important as a CPU comparison for the vague remarks I made about suspected optimization, but it does matter if you are comparing the game performance in order to make a decision about which CPU to buy if I play a particular set of games.
  • PUGSRULE! - Monday, August 10, 2015 - link

    As of today 8-10-15 an Intel i7-4960X costs $1,100 versus an AMD A10-7850k that goes for $130. If you have an unlimited budget, yeah go for the Intel. But dollarwise, you cannot do better with the AMD. It's like comparing a Hennessey Venom GT to a Mustang. Reply
  • TETRONG - Wednesday, March 19, 2014 - link

    Hmm, I agree. It's confusing why they keep reviewing and devoting so much time to something that hasn't been viable for a long time. Anyone can compare performance/price and conclude that there is no good AMD buy relative to Intels Core/Xeon lineup. It doesn't sound so bad until you factor in electricity which completely negates anything AMD might otherwise have going for them. Whatever you would save with the cheaper chip would get eaten up by the utilities -
    No thanks
  • tech6 - Wednesday, March 19, 2014 - link

    I think they can be viable at a certain price point. Offering "good enough" performance for office and non-3D gaming computers is AMDs strength and that is where OEMs should be aiming also. Making a $100+ "Extreme" AMD board makes no performance sense at all. Reply
  • bobbozzo - Wednesday, March 19, 2014 - link

    A mini-ITX Kaveri is attractive for HTPC builds, thanks to the excellent integrated graphics. Reply
  • PEJUman - Wednesday, March 19, 2014 - link


    I love my kaveri ITX build:
    1. it's quite capable for HTPC-NAS unified solution with the large numbers of SATA 3 out of A88X 2. $130 A10-7850K CPU+GPU pricing @ microcenter.
    3. $90 ASrock A88X-ITX+ @ newegg.

    Sold my ivy bridge ITX HTPC & nehalem X58 NAS. each are more capable than the Kaveri CPU. but now I run both
  • PEJUman - Wednesday, March 19, 2014 - link

    systems on 1 kaveri system alone, saving a boatload of idle power consumption + getting more capable GPU for MADVR in the process. Reply
  • duploxxx - Wednesday, March 19, 2014 - link

    not sure how you look at charts, but the issue with these reviews has always been to compare equal products, which is very obvious many tech sites can't. Again today this review is a mucked up comparison.

    Why not do a decent test with the onboard gpu and for example mantle and see the difference again, these amd series are not thrown into the market to compete on that CPU front, they are there for general purpose and mid stream market. Who ever believes he need a 4770 for general use (not all are video freaks) should think twice, but yet they can't since they are stuck in believing just benchmark results....


    309$ vs 184$
    with that price difference i buy a 128Gb latest gen SSD and you know what the AMD will fly over any application while the intel with a normal HD would cripple. so useless compare of benchmarking as if one would watch a usb3 copy being few secs faster - slower, people buy a chinese brand usb or budget usb3 device which on its own will already be slower...

    the mucked up mind is with the reviewers and believers looking purely at benches while daily almost every user is stuck with stupid MS OS.
  • Viewgamer - Wednesday, March 19, 2014 - link

    I wanted to see overclocked performance.
    It's a shame that he didn't even bother to benchmark the overclocked CPU.
    Also complaining that you get high temperatures with overclocking on the stock CPU cooler is stupid.
    Intel Haswell CPUs operate at extraordinarily high temperatures without any overclocking and yet the reviewer has the nerve to complain about Kaveri temperatures after overclocking the chip by 700mhz.
  • Ian Cutress - Wednesday, March 19, 2014 - link

    Overclocked performance is given in the overclocking section under the PovRay column in that table. Here it shows the score PovRay gets at the given overclock.

    Also, I talk repeatedly about the VRM temperatures - not the CPU temperatures. 'At stock' and 'using the stock cooler' are not interchangeable phrases. I cannot find anywhere in this review that I use the phrase 'stock cooler'. The heat given off at stock by the VRMs (a point which I highlight many times), not the CPU, can be the cause for concern, especially when the system is overclocked.

    If you would like to discuss the above issues, I do have an email you can contact.
  • Tom01 - Wednesday, March 19, 2014 - link

    There is no reason to give up AMD. The AMD FX-9590 is equal to an Intel Core i7-4770.
    That is very fast.
    I personally am an Apple-Intel user, but would prefer AMD-chips.
  • DigitalFreak - Wednesday, March 19, 2014 - link


    A 220 watt AMD process is equal to an 84 watt Intel processor?
  • Tom01 - Thursday, March 20, 2014 - link

    Yes, speed wise. Reply
  • Lucian2244 - Wednesday, March 19, 2014 - link

    As i remember this has always been a "problem" with most AMD based boards, their VRM runs hot. I guess this can be an issue in the long term but who has the time to test that :). Reply
  • alyarb - Wednesday, March 19, 2014 - link

    Recently, I paid ~$120 for a "new" Asrock board from Newegg. I received a box with no plastic wrap, no seal on the ESD bag containing the board. The socket had bent pins under the plastic cap and neither Newegg nor Asrock assumed any responsibility or offered any recourse besides a $60 repair job that would take 2 months.

    I ended up eating the $120 and bought an ASUS board that came new and undamaged. I've been dealing with Newegg since the beginning and was let down by this, so I get my LGA boards from Amazon now :(. ZIF sockets and less fragile stuff is OK to get from Newegg.
  • Hawkleberryfin - Wednesday, March 19, 2014 - link

    Why no 8350 or other AM3+ in your comparisons? Reply
  • alexruiz - Wednesday, March 19, 2014 - link

    Ian, I have a few questions from your testing:

    - What mode was the SATA controller set on initial boot to the UEFI BIOS? IDE or AHCI? In my experience, all the AsRock mobos have IDE as default. If if was indeed set to IDE, and this was a ECS or Biostar motherboard, you guys would have eaten them alive, but with the infatuation with Asrock around this forum, they get a free pass on this one.

    - Fan controls, you say you liked them. Are you referring to the graph, or the settings? If you are referring to the settings, again, in my experience, all AsRock motherboards default to full blast. Does it mean you liked all the fans at full speed as default? On a setup that very likely will be used on HTPC duties, having to go and silence every single fan in the BIOS is a NO-NO. Level 9 target as default (full blast)? Really? Again, if this had been ECS or Biostar, you would have eaten them alive for not having an easy "quiet" fan setting (Ironically, Biostar probably has the best fan settings controls, full manual control for power users looking for max coooling, and a "quiet" setting that requires only one selection in the BIOS. )

    - Hot-swap SATA AHCI. Did you have the chance to try and see if windows would recognize a SATA hard drive plugged as hot-swap? Again, in my experience, AHCI hot swap is broken in AsRock mobos, at least the AM3+ and FM2 ones.

    - Did the motherboard post normally on first power up, or did it require a CMOS clear? Again, the AsRock mobos I have used usually require a CMOS clear out of the box. What other FM2+ motherboards are under testing?

    Motherboard flex? How solid and rig did the PCB feel?

    Did you try F11 as boot override? It works, but it takes a lot of luck and several tries to catch it.

    I would like to see the FM2+ motherboards comparison once it is complete. For what I see, the infatuation with AsRock at AnandTech starts with the editors. AsRock is popular because they pack a lot of features (bunch of fan headers for example) so they give a sense of value for the money, but when the product feels unpolished, I prefer to not have the half baked feature (broken hot-swap SATA AHCI for example)
  • SolMiester - Wednesday, March 19, 2014 - link

    I just built an HTPC with the mATX extreme 4 plus A88 chipset, same as this but smaller. You are right, I did need to clear the CMOS, however Sata port were set correctly at ahci, I have no idea why you think they would be set to IDE on a board with no IDE controllers. Reply
  • hrga - Wednesday, May 28, 2014 - link

    I appreciate alex post as most of bugs did indeed be here since implementation like that BIOS boot override feature which really should have been quite a bit better especially when there's default BIOS SpalshUp wallpaper screen before disabling it. Well thats in fact with older legacy BIOS more annoying
    As for IDE option goes thats something present for all chipsets that has ability (drivers) to run under old WinXP w/o AHCI drivers for chipset streamlined. Thou i dont know who would run WinXP full time nowadays.except for some tests. Its quite a bit cleaner to capture screenshots without antialiasing artifacts even when Aero eyecandies fully disabled and put them into nitty 8b palette .png
    But those wishing to run WinXP nowadays i gues would be experienced enough to either go to BIOS to change AHCI to IDE. For a55/A75/A85/A78/A88 i believe there are no ability to run them under "stock" WinXPSP3 so IDE in that BIOS is really a nuance. And Linux kernels that support chipset couldnt pass without providing its basic features and thats AHCI. Did some prehistoric SATA-150 drives ever didnt support AHCI as it was optional in those days?
  • Ian Cutress - Monday, March 24, 2014 - link

    Hi Alex

    - ASRock have default AHCI since Z77 at least, perhaps even a little before that. I have always criticized boards in the past for not having AHCI as default, and in my testing I turn it to AHCI if it is not and explicitly point it out. So you have nothing to worry about there.

    - ASRock tends to vary their fan settings depending on the motherboard. Some of them are, as you say, stuck on full blast. This was set to 'default' by default, which gives the fan profile you see in A-Tuning. The custom multi-point gradient you can see in the BIOS also shows this.

    - I have not tested hot swap. I will have a look in future ASRock motherboards, I'm now four products further down the line in my testing.

    - I've not had issues with Motherboard flex from the big four in a couple of generations. Typically flex happens when motherboards have four PCB layers or less - anything over $100 these days tends to have at least six. But you can still build a solid four layer board.

    - I usually go into the BIOS for boot override so I can double check the settings as I go through. It terms of hitting the button at the right time, it usually depends on when the USB driver from the BIOS is implemented. Best way to help yourself in this is to put the keyboard in a USB 2.0 slot, those are initiated first. Some motherboards are better for this than others, but you have as much chance on the same board of getting into the BIOS as boot override.

    Each of the motherboard manufacturers have their quirks. ASRock likes to try a lot of little things each generation and see what sticks - some of these work well, and some of these do not. They have been amenable to suggestions over the 3+ years I have been at AnandTech, and they are taking some of them on board. There is no infatuation here - if I could look at these motherboards in a double blind study I would do, but enough of the product is personalised that that is not possible. I aim to look at each motherboard afresh, and sometimes there are features on motherboards that don't make sense for 99% of people that inflate cost. Motherboard manufacturers also have to guess a lot of the hardware specifications 8-12 months ahead of launch, and it can be hard to get them right and still remain competitive (and get it all to work). Sometimes each of the companies do come out with some bad stuff, and sometimes they come out with the goods. Awards are well deserved and should be for the best, hence why at AnandTech we rarely give them out compared to some others that have an award fetish because it helps promote their site/get them more review samples. We are lucky enough not to be in that position, and potentially help direct evolution of product, for any company that wants to listen.
  • Bob Todd - Wednesday, March 19, 2014 - link

    Pretty nice feature set for the price, especially before any sales or rebates. However the comparatively dreadful CPU performance still makes the market look pretty small. Excellent for HTPC usage or that small subset that wants to do medium settings 720p gaming without a discrete card. 8x 6Gbps is quite nice, but I don't imagine many people going for midrange CPU performance are going to buy oodles of SSDs to fill all of those (mechanical disks or gigabit limited NAS setups won't really see the benefit). Getting your ass kicked by a 3 generation old 2500K that was in the same realm pricing wise is nothing to get excited about. The once meaningful 'budget' advantage mostly disappeared way back with the SB Celeron chips. And with Bay Trail Intel ended the one mainstream segment that AMD was kicking their ass in since Bobcat (yes AMD has the better iGPU, but the CPU isn't fast enough to do anything meaningful with it). And I realize that was mostly an off-topic rambling since this article is about the board not the CPU. I just miss you AMD. Please be competitive again. Reply
  • popej - Wednesday, March 19, 2014 - link

    "Despite the Realtek ALC1150 having a 115dB signal-to-noise rating, our RMAA test gave a result nearer 100".
    Do you mean dynamic range measurement in RMAA? According to ALC1150 datasheet, dynamic range is 104dB.
  • Elmstreet - Thursday, March 20, 2014 - link

    Just so I get this correct, most of the gaming benchmarks are comparing the Intel i7-4960x, which currently sells for over $1,000 on NewEgg, with an AMD A10-7850k that is selling for just under $200. Based upon my disposable income, this is not a fair comparison at all. This is highly biased towards Intel.
    If it was to be a proper comparison, the Intel chip would also need to be in the $200 range. Perhaps an i5-4570?
    I'm not trying to bash on Intel at all. I agree with almost every single post on here about how AMD has been failing for such a long time, and that Intel has taken over. But, come on. At least make the fight balanced.
  • fteoath64 - Friday, March 21, 2014 - link

    True that comparison between Intel and AMD always has a huge price difference. It would be better to compare IGP to IGP and discrete gpu with discrete gpu of the same price range of system. Much like TomsHardware did in terms of setting up a budget and getting the best components for the budget, then optimise it a little like +$100 gives this!. + $200 gives these choices with these performance figures. This in effect, people tend to buy AMD for their gpu and prices than anything else. IF they are into discrete cards, tendency is for Intel cpus.
    Other reason to get Kaveri is HSA, of course for those in the know. But the drivers and OpenCL 2.0 is not available as yet. So, question would be , next version of Kaveri might have bugfixes and optimizations to really boost HSA functions, or might not (as in Richland from Trinity jump). People tends to compare 2 or 3 key components that are "must haves" like 128GB SSD, then discrete gpu or decide only iGPU graphics.
  • Ian Cutress - Monday, March 24, 2014 - link

    This is a review of the motherboard after all, designed to compare against other motherboards of this type as we review them.

    This is also my new 2014 benchmark suite, implemented for Kaveri launch. If every time I updated the suite I tested a bunch of processors for the first review, we wouldn't get any reviews of anything for a few months (and no-one gets paid). So we are testing as we go, as is usually the case. I plan to do an update to my Gaming CPU articles when I get the chance to test a few more CPUs. At the minute the number of motherboards I have in to test is shockingly long.
  • 5thaccount - Thursday, March 20, 2014 - link

    I know this isn't a review of the APU, but I'm very curious why it wasn't benched against a Celeron / i3 / i5 in the system and game benchmarks. I wouldn't expect this to come near i7 / Xeon territory (although, to it's credit, it did better than I was expecting). Was it only to strictly compare relative performance to the fastest Intel can provide? I'd be curious to see how it fared against similarly priced and cheaper Intel CPUs - especially in the game benchmarks. Reply
  • aike - Friday, March 21, 2014 - link

    All the shock practice on the devices can be used keep on effort just by advantage about keeping effort or possibly at home process. All the devices undeniably construct over to any <a href="">rolex replica</a> private??? lenses temperament, considering that it adds something to a?? lenses is visually and additionally trend. Reply
  • aike - Friday, March 21, 2014 - link

    Genuinely, each one of on top of exploit fundamental purpose typically the copy wholesale handbags turned out to be some type of shrewd undermine concerning trendy finishing touches an increasing number of gals get to approve perhaps even booty. Reply
  • DrMrLordX - Saturday, March 22, 2014 - link

    Thank you for taking the time to review this board, Mr. Cutress. I am curious about CPU throttling under heavy IGP load. Did you experience this problem? Some have attributed this to VRM temperatures, but others have observed that this may be related to a design decision on AMD's part to keep total power delivery to the chip within a specific TDP envelope, even when overclocking is present.

    I have been trying to figure out which AM2+ boards can mitigate or defeat this throttling behavior through UEFI settings. There is another way not involving the UEFI:

    but it would be much nicer to tweak the UEFI to defeat this p5 state behavior. People with "good" cooling on the CPU and VRMs are still getting throttling, at least on some boards, some of the time . . . it's hard to nail down exactly who has this problem.

Log in

Don't have an account? Sign up now