Updated: Retail Radeon R9 290X cards may be slower than press samples

Modern CPUs and GPUs employ turbo mechanisms that adjust frequencies based on various factors. Although they aspire to higher speeds, these clock-boosting schemes have well-defined base frequencies that guarantee a minimum level of performance. Or most of them do, anyway. AMD defied convention with its Hawaii-based Radeon R9 290 and 290X by only publishing the peak burst frequency for those cards. We know the GPU speeds under ideal conditions, but there’s no minimum baseline.

Our reviews of the Radeon R9 290 and 290X illustrate that both cards have trouble maintaining their peak boost frequencies while running games. As the GPUs heat up, the clocks scale back from their peaks, and frame rates drop accordingly. The performance hit only works out to a few FPS on the press samples we’ve tested. However, Tom’s Hardware has seen a bigger performance drop on a couple of retail 290X cards purchased from Newegg.

According to the site, the GPU frequency of one of the cards falls as low as 727MHz—well below its 1GHz boost rate—and stays there. As a result, the retail 290X is notably slower than not only the equivalent 290X press sample, but also the Radeon R9 290 supplied by AMD. Tom’s doesn’t provide much detail about the behavior of the second retail 290X, but it says that card is also slower than the 290 press sample.

AMD told Tom’s Hardware that something is wrong with the retail cards, and I’d tend to agree. If the GPU consistently pulls up 27% short of the advertised boost speed, that ain’t right. But since AMD hasn’t defined a base frequency for the R9 290X, a 727MHz sustained GPU speed technically isn’t out of spec. We’ve asked AMD to comment on this matter. We’ve also inquired about whether the 290X should be running as slowly as 727MHz under normal gaming workloads.

Chip-to-chip variance is common in semiconductor manufacturing. Any overclocker can tell you that some GPUs are simply more comfortable running at higher speeds—and with lower voltages—than their peers. Those differences typically don’t produce large performance deltas between stock-clocked products that share the same model number, but that appears to be what’s happening with the Radeon R9 290X. The dynamic PowerTune mechanism in the Hawaii GPU, coupled with AMD’s apparent desire to wring every last drop of performance from its new Radeons, seems to have created a situation where performance can vary quite a bit based on the individual characteristics of each chip. Worse, it’s possible reviewers may have been seeded with cherry-picked samples capable of maintaining higher speeds than typical retail products.

We’re eager to hear more from AMD about this issue, and we’ll update this story as new details roll in. We also have a couple more review samples we can test internally. If you bought one of the new, Hawaii-based Radeons, you can monitor the GPU frequency using GPU-Z, which is available here. We’re curious to hear your results.

Update: AMD has issued a statement on the matter.

A media outlet has uniquely reported instances of AMD Radeon R9 290X boards purchased in retail that have exhibited an uncharacteristic level of performance variance as compared to press samples issued by AMD. We’re working to secure the board(s) in question for further analysis.  Boards purchased by other media outlets have not exhibited similar characteristics that we’re aware of. In the meantime, we’ve identified areas where variability can be minimized and are working on a driver update which will minimize this variance.  We will provide an update shortly.

Interesting. We should note that TR reader JohnC has encountered clock speeds well below 1GHz on his retail 290X. You can read about his experiences in the forums.

Comments closed
    • sschaem
    • 6 years ago

    JohnC graph show 900mhz average over 50 minutes (stock)
    Thats actually really good… The issue reported is average of ~720mhz
    So his card does not show the problem reported.

    AMD seem to have a few problems with fan speed, but also the nature of the beast. (asic quality variance)

    [url<]http://www.tomshardware.com/reviews/radeon-r9-290-driver-fix,3666-4.html[/url<] Also, because the card are literally dynamically 'overclocking' (thermally) to asic max, not all card will be the same. All card can go to 1GHZ, if proper cooling is provided. and the stock cooler cannot do it. The higher the voltage, the more of a wall you hit. Moral... wait for the dual fan OEM models if you want to be sure you dont hit the thermal wall.

    • allreadydead
    • 6 years ago

    Well, at least AMD learned from a thing or two from past;

    “If it can’t stand the heat, take it easy and don’t keep heatin’ it ’till it’s toasted”

    • alienstorexxx
    • 6 years ago

    nice way to harm amd’s image. journalists should apologize for their incompetence sometimes.
    figure this title imact on a future amd user.
    “Retail Radeon R9 290X cards may be slower than press samples”
    if the card fan was running on lower speeds, th should figure it before uploading the review, or at least make some type of checking.
    can’t believe how easily people draw lot of conclusions about this and everyone involved amd as the mainly responsible.
    “golden samples”

      • Airmantharp
      • 6 years ago

      If that was the intent, then it was very well executed; the accusation is entirely in line with the card’s performance characteristics, and this is exactly what we expected to happen if a card has a less than stellar GPU that runs a little hot or if the card is placed in a stifling environment. And that’s not even going into the possibility of poor assembly.

      Remember that TR mentioned how even minor changes in environmental variables resulted in performance adjustments by the card itself, given that it was running at it’s thermal limit when loaded and could not increase it’s fan speed, then it must reduce it’s power draw to compensate, and voltage and clockspeed must be curtailed to accomplish that.

        • alienstorexxx
        • 6 years ago

        that perfectly adresses to answer my post…

          • Airmantharp
          • 6 years ago

          You’re welcome 🙂

    • ptsant
    • 6 years ago

    It appears that the specific retail card had a defective fan, according to tomshardware.de (if I understand correctly). The same card would get normal performance levels when the fan was spun as fast as on the press sample. For more info see:
    [url< ]http://www.tomshardware.de/radeon-r9-290-golden-sample,testberichte-241424-2.html[/url<] Furthermore, other users (http://forums.overclockers.co.uk/showthread.php?p=25255073#post25255073) have tested retail cards and have found no deviation from press results.

    I have to assume that this card is pushed very hard for its particular cooling and any minor change in thermal performance is immediately visible in benchmark results. The answer is not complicated: just wait for custom cooling solutions, pay $10-20 more and get superior performance and peace of mind.

    • NeoForever
    • 6 years ago

    In an AMA session today at Tom’s, AMD claimed that TR and THG has faulty boards. Will TR edit / repost the review after AMD releases the drivers that fixes these boards today or tomorrow?

    [url<]http://www.tomshardware.com/forum/id-1863987/official-amd-radeon-representatives/page-5.html#11884668[/url<] The post is long, try looking for "Plain and simple".

      • clone
      • 6 years ago

      interesting and nice to see some involvement, Warsam71 hasn’t managed the same.

      • Airmantharp
      • 6 years ago

      Since I won’t be visiting THG, just how exactly did TR get a faulty board? You think they’d of noticed?

        • clone
        • 6 years ago

        given it’s a driver tweak away from being fixed I’m more convinced it’s an early release bios.

        especially given their is only one board maker and they are all going down the same line.

        • spuppy
        • 6 years ago

        He was referring to the TR forum poster, not TR themselves

        At least the one asking the question was:

        “and even a user on Tech Report claiming much lower clocks than review samples”

        answer:

        “Plain and simple, THG and Tech Report have faulty boards”

        That actually makes it more interesting, because they are dismissing customer reports of slower speeds as faulty boards.

    • Fighterpilot
    • 6 years ago

    I’m amazed that you would still have a “story” like this on the front page of TR.
    <crickets>
    #not worthy

    • tbone8ty
    • 6 years ago

    Damage labs should slap an aftermarket cooler on an R9 290x and retest!!!

    🙂

    would be interesting results!

    The 280x aftermarket coolers fit.

    cant wait to see an asus or msi cooled 290

    edit: [url<]http://www.techspot.com/review/736-amd-radeon-r9-290/page8.html[/url<] [quote<]"Update: Based on your feedback, I took the IceQ X2 cooler off the HIS Radeon R9 280X and stuck it on our R9 290 sample. Cooling was dramatically improved. The FurMark stress test maxed out at [b<]76[/b<] degrees while the card never exceeded [b<]63[/b<] degrees in Crysis 3 and Battlefield 4. So it seems as expected the board partners will be able to solve the heat issues of the reference card." [/quote<]

      • Airmantharp
      • 6 years ago

      Damn, I was hoping that we were looking at a transplant of HIS’s blower; certainly, though, open air coolers will not be challenged by these cards, and that’s great to see!

    • TwoEars
    • 6 years ago

    The latest rumor I’ve heard is that there MIGHT be a slight performance difference between different samples when using the stock AMD profile and not the so called “uber” profile.

    I believe AMD is going to issue a new bios version, with a slightly more aggressive stock fan profile, to mitigate any variations that there are between individual samples.

    To the best of my knowledge no one has yet reported any performance difference when using the “uber” profile. It’s only when using the “stock” profile that some user have cards that run slightly hotter than others and therefore go into thermal throttling sooner.

    In my mind this is all a result of AMD wanting to maximize the performance of their card and wanting to bring as much bang per dollar as absolutely possible to the end user.

    For that I applaud them.

      • Airmantharp
      • 6 years ago

      What really seems to have happened is that AMD built a very, very nice card with a very nice blower. You don’t get to over 50dBA without driving people out of the room using a crappy, whiny blower, and many reviewers noted that the blower isn’t as offensive as the SPL readings might suggest, though it is indeed noticeably loud.

      Thing is, they decided at some point after committing to the board and cooler design that they needed to get more stock performance out of it, and pushed it to it’s absolute limit. It’s the same thing that we’d expect from any card if over-volted and over-clocked while using increased fan speeds to keep thermals under control.

      So there’s definitely a value argument here, especially considering that third-party coolers will make quick work of the heat these cards generate while keeping noise well under control. If evaluated under that light, AMD rightly deserves the near-universal applause that’s been directed their way.

    • Dezeer
    • 6 years ago

    So apparently the software fix is just going to be normalizing the fan levels to same rpm level instead of same percentage level.

    • DPete27
    • 6 years ago

    In related news, [url=http://www.tomshardware.com/reviews/radeon-r9-280x-third-party-round-up,3655-5.html<]Toms also reviewed a variety of 280X retail cards with custom coolers.[/url<] That's a pretty good indication of what you can expect from the R9-290's with custom coolers. They fare much better than the [url=http://www.tomshardware.com/reviews/radeon-r9-290x-hawaii-review,3650-30.html<]AMD reference blower[/url<] in terms of noise AND temperature.

    • LukeCWM
    • 6 years ago

    The two news sites I go to are Tom’s Hardware, for a quick gist of many stories (that are poorly written and often sensationalist), and then TechReport for quality reporting on fewer topics from people I trust (and a great comment community to boot!)

    I saw Tom’s had an article on the 290 hours before TR posted about it, but I didn’t even open it. I’m so used to only trusting TR for graphics reviews, I guess I’m in the habit of not even opening them at Tom’s anymore!

    • Bensam123
    • 6 years ago

    Curious what ‘areas of variability’ are… That doesn’t seem like simply ramping up the fan speed more.

    It is rather ridiculous how much hate this is getting now. At the very least they have to tweak the fan setting by 5% and people are in a outrage over it, completely putting aside the fact they released a $550 card that pulled Nvidia offerings down by like 50% and then a new one that will pull them down even further to the buyable levels for mere mortals.

    Hopefully AMD will come out with a second gen cooler shortly if this is purely a heatsink issue. If it isn’t, everyone wins… I’d like to know what they need to tweak to reduce the heat buildup… VRM phase issue or something?

    • superjawes
    • 6 years ago

    Even with the update from AMD, it would still be nice to see an independent analysis of these cards. Like I said before, the performance benchmarks don’t mean anything if the performance tanks in real world systems.

    • deinabog
    • 6 years ago

    Hmm, I guess I’ll stick with my GTX 770s then (not that was considering R290X anyway). It is unfortunate that AMD’s latest and greatest is having this issue when just recently many tech sites were singing its praises. At times it seems the company can’t catch a break.

      • Airmantharp
      • 6 years ago

      They are the masters of their own fate… every issue that’s been raised with these cards was a poiont of concern well before they were released.

        • clone
        • 6 years ago

        I suspect the problem is poor case cooling, poor flow, closed enclosures, with the question on wether or not it’s deliberate left up in the air. I find it impossible to believe that AMD released a card to the press that’s different from retail samples.

        not in the age of the internet where information moves this fast and everyone has been well aware of it for the past 20 years….. those in the know really aren’t that stupid and things are never that simple.

        a hatchet job can be manufactured far more easily by a few than can a reputation be cleared, while I can’t speak to this being the case in this instance doing it would require no effort.

        p.s. the claims have already been debunked.

          • Airmantharp
          • 6 years ago

          Yeah, I read about the debunking right after I read about the ‘issue’, and as it was THG, I had/have the same suspicions regarding the testing environment and the integrity of the testers, and I have far more faith in AMD.

          And really, this is exactly what we expected to happen- the card runs as fast as it’s thermal limiter will let it. As most reviewers and owners have mentioned, any change in the thermal variables directly affects the card’s performance, so we should expect throttling in less-than-ideal setups.

          • Diplomacy42
          • 6 years ago

          Hmm, I think that YMMV and Caveat Emptor apply to this situation. if EVERY reviewer and pre-tester says the card runs hot and loud, and then you buy it, and it runs a little hotter or a little louder, than you expected, then you chalk it up to process tolerances and not a bait and switch conspiracy.

          now if nobody had noticed the mysterious heating/cooling issue that popped up ONLY in this batch of cards I could see possibly a modicum of “outrage” although it isn’t really uncommon for the media version of the card to have subtle differences compared with the final version. It would still be quite the leap to intentionally misleading people.

          • Bensam123
          • 6 years ago

          Aye, it’s extremely easy for people to figure out when samples are cherry picked and the kind of reputation people currently regard AMD with, they wouldn’t pull a stunt like that. All I’ve seen come out of AMD the last few years are good intentions for the most part. Open source programs, open initiatives, lowering price/performance baselines, building new foundations to base future technology off of for the betterment of everyone (GCN, HSA, Trueaudio, Mantle). They’ve actively attacked problems people presented them with (crossfire stuttering) and met their goals with them while still working on problems others have pointed out (4k issues).

          People need to cut AMD some slack, especially considering their financial situation they’re doing amazingly well. If they can still bring a giant down to their level (Nvidia) they’re still fighting.

    • Tristan
    • 6 years ago

    No wonder, why Origin PC dumped this ‘graphics card’. But AMD’s fanboys and trolls, of course always blame NV and money.

      • Spunjji
      • 6 years ago

      Yes, their announcement which took place before this card was sampling was clearly referring to it. The fact that other organisations were approached to perform the same stunt is immaterial.

      /sarcasm

        • HisDivineOrder
        • 6 years ago

        Link for “other organizations were approached to perform the same stunt?”

    • Yeats
    • 6 years ago

    A user at the AT forum is saying that his card’s fluctuating speed appears to be due to V-sync.

    [url<]http://forums.anandtech.com/showthread.php?t=2351689[/url<]

      • Airmantharp
      • 6 years ago

      Well, that should cause fluctuating speeds, since V-Sync along with AMD’s frame-pacing technology would probably result in idling the GPU and thus decreases in thermal constraints for bursts of time.

    • TwoEars
    • 6 years ago

    A Swedish magazine already debunked this. They bought over the counter samples and tested against review samples and they were just a fast. Absolutely no difference.

    More likely is that whoever got a bad results didn’t have enough cooling and that the card went into thermal throttling.

      • Blink
      • 6 years ago

      Link please. Search engines are failing me.

        • MikkleThePickle
        • 6 years ago

        [url<]http://www.sweclockers.com/artikel/17845-snabbtest-radeon-r9-290x-i-blasvader-for-golden-samples-sweclockers-undersoker[/url<]

          • Blink
          • 6 years ago

          Thank you. The result is what I expected. However I do hope some other review sites can duplicate the results.

            • MikkleThePickle
            • 6 years ago

            Anytime 🙂 I also hope we get more information from other sources soon.

          • Goty
          • 6 years ago

          I bet that doesn’t get added to the story =P.

          • LukeCWM
          • 6 years ago

          Just reading the link made me laugh out loud! So Swenglish! =D

      • puppetworx
      • 6 years ago

      I’m actually disappointed TR published this instead of waiting for or producing corroborating data. It’s one thing to report on leaks without corroboration but when the thing being reported can be [b<]easily[/b<] checked and verified because the products exist in the public domain it seems crazy to publish without any corroboration. I reserved comment on this story until I saw someone else examine the issue precisely because I didn't believe I could trust THG solely. It now seems I was right to do so. I suppose one school of thought is that it is in the TR reader's interest to know about a potential problem with a new product. In my opinion reporting this before gathering supporting or opposing evidence was poor journalism, whatever the result turns out to be. The underlying accusation here is fraud, it's not the most extreme form of it but that's what it is: AMD knowingly sent parts to reviewers which were different to what the consumer could buy in the market but which were branded the same. That's a serious charge which should heed some investigation, especially in the absence of corroboration, before reporting it.

        • Airmantharp
        • 6 years ago

        What do you think the best way to get community feedback on an issue like this is?

        I think TR succeeded quite well; the community is digging into it, and AMD’s attention has been paid. This is how you get things fixed, and this is how people with sway in the industry take care of the community around it.

          • puppetworx
          • 6 years ago

          The community doesn’t have access to cards given to the press. How exactly are they supposed to compare them against cards from the marketplace? The only people that can do that is the press.

            • Airmantharp
            • 6 years ago

            By comparing their results, especially if it’s not what they’re expecting to see? There are some talented people out there, beating the hell out of these cards right now for this purpose and that. Any substantial issues will surface pretty quick.

            • puppetworx
            • 6 years ago

            Ludicrous. The only scientific way to compare press cards to market cards is by comparing them both on a standardized test bed, doing what you suggest introduces variables which render the data useless.

            Anyone in the press could have knocked out a scientific test in 30 mins to corroborate or refute THG’s observation. I’m at a loss to why hardly anyone did.

            • Airmantharp
            • 6 years ago

            I think you’ve answered your own question- no one could. Thing is, we didn’t need to test them side-by-side, just identify that there might be a difference to justify the testing, and that’s been done.

            • puppetworx
            • 6 years ago

            It was my original contention that the community could add nothing, only the press could. You’re saying THG identified what seemed to be a discrepancy, which ‘the community’ had no means of doing or verifying, but it was ‘the community’ which proved that further testing was required? No, it was THG that proved that and it should have been tested for corroboration before being reported on.

      • Diplomacy42
      • 6 years ago

      well, we already knew that the cooling solution on the card was woefully inadequate. Its not surprising to see anecdotal accounts of throttling in borderline rigs that haven’t been causing problems in the past.

      • Bensam123
      • 6 years ago

      Don’t tell JohnC…

        • chuckula
        • 6 years ago

        You know Bensam, for an AMD fanboy you sure seem to spew a lot of hatred towards a guy (JohnC) who actually bought the product that you worship so religiously but didn’t actually bother to purchase yourself.

        Then you try to crucify him when he makes some honest and factual observations about how his particular card behaves, followed by a bunch of sanctimonious drivel like “Oh, just because you actually own the card and have made repeated observations to how it behaves with multiple fan profiles, you’re an idiot because you don’t automatically conform to my preconceived notions of how I, Bensam123, God among Insects, WANT the card to behave!”

        Sorry Bensam, I’m going to take JohnC’s word over yours every time until YOU can post REAL RESULTS from YOUR OWN CARD instead of insulting people who have done a whole heck of a lot more to keep AMD in business than you have.

          • Bensam123
          • 6 years ago

          This is the thread he’s talking about:

          [url<]https://techreport.com/forums/viewtopic.php?f=3&t=90345[/url<] JohnC said his airflow intakes were taped up and he was using BF4 to test things. He hadn't tested the above with the side of the case off to weed out the intakes and he wasn't testing with a easily repeatable workload (Furmark). The intakes in particular are of concern when you're talking about cooling performance and I would rightly again suggest the same thing. You need to weed out confounding variables in order to make a test scenario easily reproducible and reliable. That is part of the basics of the scientific method. If you scroll halfway down the page you can see my 'crucifixion' or you can read this: [quote<]If you're having problems with negative pressure just increase the air intake, you can do that with slot covers on the back of the PC or simply getting better fans for intake. You don't need to switch to a recirculating fan. Speaking of negative air pressure. Have you tried running the card at default settings with the side of the case off? Are you sure you don't have negative air pressure? You said you taped up all the intakes besides the filtered fans... I'm guessing the intakes were part of Antecs design. The Titan cooler recirculates more air then AMDs does based on the design of the cooler (half the air goes back into the case). Curiously it almost seems like you're boasting about how awesome your Titan is when compared to the new 290x you just bought... I don't know why people aren't running Furmark to see what kind of frequency they get with that. Instead they're trying to hodge podge it with a game.[/quote<] Worship? YOU are the one that says I'm a AMD fanboi. I think this is another case of just how untrue that is. Spewing hatred? That's what you just did. Crucified? Like asking someone for further testing on something and getting slammed by it in personal attacks (your post) and by a mod that somehow overlooked the fact that there was needed clarification on his testing. Edited out due to a misunderstanding of the original ban. Chuckula still acts like a douchebag and this has been going on for close to a year.

      • Pholostan
      • 6 years ago

      The article was updated yesterday with official word form AMD:
      [quote<]We will be releasing a driver in the next 24 hours that corrects this behavior by normalizing all fan behavior. The 290X in Quiet mode should be at 2200RPM, and the 290 should be at 2650RPM.[/quote<]

        • Bensam123
        • 6 years ago

        So they fixed it by increasing fan speeds or fixing fans that weren’t running at the right speed?

        Perhaps this is caused by variance in the fans themselves (+/- 10% RPMs). Originally they based fan speed on load rather then on RPMs…

    • internetsandman
    • 6 years ago

    From what I’ve seen, the 290X can deliver mind boggling performance, IF you’re willing to drop a bit of extra coin and time and put it into a watercooling loop that takes heat out of the equation. I’d really love to see a review of a 290X that’s been fitted to a decent WC loop, and see what happens to the boost speeds once it’s given enough thermal headroom to stretch its legs a bit. I don’t know how practical it would be for a reviewer to do that, but I think it’d be interesting to investigate a best-case scenario on a card that clearly has potential

      • Diplomacy42
      • 6 years ago

      define “mind boggling”

        • HisDivineOrder
        • 6 years ago

        Mind boggling is defined as that which boggles the mind.

        …You’re welcome.

          • Diplomacy42
          • 6 years ago

          its bad form to use the word in question within its own definition.

    • Mat3
    • 6 years ago

    Nvidia’s been doing this a lot longer than AMD. They have their minimum speeds but you can be sure the review samples are cherry picked for the best boost.

    • Bensam123
    • 6 years ago

    And then the hate train starts again with a single card from THG… I’m guessing if it’s really that bad they’ll issue a fan speed bump. I don’t know why it doesn’t increase the fan speed anyway when the core speed drops below a certain threshold (like 900-950mhz) instead of having a static maximum for the fan. That seems like the obvious choice to take care of outliers if normal performance does reside within +/-10% normally.

    People are less likely to complain about noise rather then performance.

      • Bensam123
      • 6 years ago

      Yay, I got -6 for suggesting having a dynamic fan speed for all cards!

        • Airmantharp
        • 6 years ago

        I’d think, with noise at these levels, people are actually more likely to complain about noise rather than performance; which do you think they’ll notice first? We know what AMD thinks :).

        Anyone willing to tinker even slightly can remove the fan limiter and let the card run free, of course. It seems that AMD picked the lesser of evils.

    • NeoForever
    • 6 years ago

    I think this story was leaked by Snowden-from-the-future but he made some big typos.

    Instead of 290X retail samples, it’s supposed say 780Ti 😉

      • Meadows
      • 6 years ago

      Yeah, right.

      • NeoForever
      • 6 years ago

      lol at least 13 people thought I was serious or are nvidiots

        • zaeric19
        • 6 years ago

        At least 14 people think posts such as your original add no value to the article and fail to create a meaningful discussion. Thusly, it would be better if it was omitted.

          • superjawes
          • 6 years ago

          15…and probably rising.

          • NeoForever
          • 6 years ago

          My comment (unlike some others) must not be funny or sarcastic enough to qualify for adding “value” or creating “meaningful discussion”.

          I will craft my (un)witty comments more carefully next time. :lols:

            • Airmantharp
            • 6 years ago

            Read up on Count Chukula, he’s the master 🙂

      • kamikaziechameleon
      • 6 years ago

      That gave me a good laugh.

    • USAFTW
    • 6 years ago

    Pity…
    A good GPU… ruined by the stupid cooler.
    If it could keep the thing at constantly lower temps this wouldn’t happen. Still, nothing that can’t be fixed with a simple driver update that lets the fan spin up to over 9000!

      • Prestige Worldwide
      • 6 years ago

      I think at $399, tacking on another $100 for a full-card waterblock gets you a very nice, cool, and quiet card in Titan territory.

      I’m definitely tempted but am hoping for the 780 to drop down in the $350 range on some kind of epic black friday sale and get that instead.

        • Deanjo
        • 6 years ago

        [quote<]another $100 for a full-card waterblock[/quote<] Assuming that you already have a full blown water cooling system that is up to the task of handling the additional heat from the 290X. If not, the cost is a lot higher.

          • Airmantharp
          • 6 years ago

          And having to pay for, and install, all that equipment makes the price delta for a good Nvidia blower look like a bargain, assuming you have to buy a card today :D.

        • Diplomacy42
        • 6 years ago

        I’ve been interested in water cooling for a long time. Until now, I couldn’t justify it.

        Now, lets be honest, I STILL can’t justify it, but I want to 0.o I really want to.

          • Airmantharp
          • 6 years ago

          Same here- I really wish I we had more data on G-Sync monitors. Us older enthusiasts have bought it hookline and sinker :).

    • Krogoth
    • 6 years ago

    This has been going on for a while on both camps with the whole turbo and power boosting business.

    GPUs themselves don’t use all of their power if they are CPU limited in an application and power profile throws the GPU back a notch or two as a result. However, you can override this with some third-party tools at your own risk. (mostly done by overclockers).

    I’m not surprised that it is more noticeable with 290 and 290x since they are getting close to their thermal limits and it all it takes is having less than ideal airflow conditions to cause throttling problems. The same thing can happen with higher-end Nvidia GPUs if you neglect thermal management in your chassis. I do remember throttling being a big problem back with the higher-end Prescott CPUs and first-generation Phenom CPUs.

      • HisDivineOrder
      • 6 years ago

      The problem is when a product promises only “up to” a certain speed rather than giving a definitive “base clock” that then boosts up to a higher number that may or may not happen, depending on conditions.

      AMD doesn’t give you the base clock at all. I think that’s where the problem is. They’re afraid to list the real base clock because it’s rather low.

        • Airmantharp
        • 6 years ago

        I don’t like posting links to other site’s reviews, but the ‘Core Clock’ speeds are listed by [url=http://www.anandtech.com/show/7481/the-amd-radeon-r9-290-review<]Anandtech's R9 290 Review.[/url<] And it is surprisingly low, compared to other card's 'base' clockspeeds.

    • DPete27
    • 6 years ago

    Isn’t AMD shooting themselves in the foot by strapping junk coolers to these cards? Shaving $5 from the reference BOM is definitely not worth the dismal thermals/throttling that’s being reported by reviewers. Gives the company a bad rap.

    Also, except at initial launch, [i<]most[/i<] retail GPUs will have a modified cooler anyway. Those coolers typically come with small price premiums at launch compared to the reference, but because of the scarcity of reference cards, that price premium is the new "norm." Those modified/aftermarket retail coolers almost always outperform the reference design. (except in the case of the "Titan-style blowers" on the high-end Nvidia cards)

      • bfar
      • 6 years ago

      I think folks would happily pay an extra 30 bucks for a good cooler. As expensive as the 780 was, the quality of the product was reassuring. It’s unfortunate that AMD didn’t step up to the mark in this area. Even if this story is BS, it will probably stick. A number of reviewers including Anandtech have complained about the stock cooler.

      My personal feeling is that I could probably live with the heat and noise at that price point, but I’d be nervous about putting two of those cards beside each other in xfire.

      For producing top drawer performance and quality at a fair price, AMD have clearly built up well deserved popular support with the enthusiast community. It is such a shame to see that scratched, especially when it could have been easily avoided.

        • pixel_junkie
        • 6 years ago

        FWIW, Guru3d tested the 290 in a Crossfire config. According to what they wrote the temps were the same (no big surprise there since they already run at max temp when alone) and the noise level was only 2dB louder in Crossfire.

      • juzz86
      • 6 years ago

      In all honesty, it’s not a horrible move by AMD. The reviews are out, and they show comparatively good performance versus the competition, with a typical horrible-as-all-buggery AMD blower on top.

      So the reviews paint the picture that performance is good, and gets us hankering for aftermarket versions, because ‘if it goes that well with that PoS on top, imagine how much better it’ll go with an ACX or a VaporX!’

      Sure AMD aren’t going to sell a boatload of reference boards. But they pave the way for the retail partners to clean up 🙂

        • Firestarter
        • 6 years ago

        [quote<]imagine how much better it'll go with an ACX or a VaporX![/quote<] well I think you'd be bumping into the limits of the power circuitry with those coolers, long before the cards actually hit the target 95C edit: disregard what I said, the cards obviously run at ~1000mhz when they're not overheating, so any big enough cooler will keep them at that speed constantly. I was thinking of the boost schemes as previously implemented by AMD and Nvidia.

      • Modivated1
      • 6 years ago

      I hope a retail partner can produce a quality cooler to contain temps, I hadn’t doubt’ed that it could be done but now I am a little worried.

      Hopefully like AMD is claiming, this card situation is more or less isolated for the dramatic drop off. I think Tom’s Hardware would have reported in more detail concerning the dramatic deficiency to maintain top clocks on both cards.

      Needless to say I will wait until a good amount of aftermarket third party card reviews are out to decide whether this problem is a persistent one. Only time will tell if this card is hyped for performance it can’t maintain.

      That would be a sad day for me, even sadder for AMD-would be burn’in the bridges of Customer Respect.

      • l33t-g4m3r
      • 6 years ago

      AMD’s never had a great blower. They all under perform, and are overly loud. I won’t touch non-blower aftermarket cards either since a lot of those are junk, plus I prefer the heat exhaust capability of a blower. Blowers are sturdier, being anchored to the pcb, while cheap aftermarket cards are often not.

      If AMD could match nvidia’s titan cooler, then I might consider a purchase. The drivers need to be better too.

      • HisDivineOrder
      • 6 years ago

      I don’t think they are shooting themselves in the foot.

      Look at what people are saying. There are LOTS of people who are deaf, who haven’t had their near miss with a loud card, or who just love AMD that much (or hate nVidia that much) that they don’t CARE. They’re willing to get on board the ear-damage brigade and ride on to Hawaii.

      Nothing wrong with that, but I know I won’t be getting on that boat. Still, AMD would have spent money developing something new they didn’t have to. Anyone who looks at the card to complain has to basically say, “Well, we can’t be sure how it will end up because we don’t know what the custom boards will be like.”

      So the fans will ignore the problem, the critics can’t reasonably argue against the card yet at launch or weeks, even a month into launch, because the “final versions” aren’t really available, and so like with the 7970 GHZ card reviews, you’re going to have months and months of people arguing that anyone quoting a review of the 290(x) is just talking about the reference and that doesn’t really apply to the custom cards at all.

      This creates uncertainty and people love uncertainty when it’s around something they want to be great and can’t stand if it’s not. They can use the uncertainty to believe… whatever they like. Meanwhile, the critics who won’t like anything AMD no matter what will do the same… in reverse. They too will believe… whatever they like.

      The reference boards are like clouds floating by. One guy sees a redhead riding a hellhound straight from hell up a shocked and terrified fairy’s butt. His buddy sees a redhead on fire, burning while a fairy stands nearby laughing.

      That’s what having reference boards that aren’t very representative of much of anything gets you.

      • Arclight
      • 6 years ago

      Idk if it’s $5 since the card would need a cooler simillar to the HD 7990, but I agree, they should have made that investment to avoid all this drama. A $600 R9 290X and $450 R9 290 would have been looking a lot better with one of these coolers:

      [url<]https://techreport.com/review/24703/amd-radeon-hd-7990-graphics-card-reviewed[/url<]

      • Arclight
      • 6 years ago

      [quote=””<]Those modified/aftermarket retail coolers almost always outperform the reference design. (except in the case of the "Titan-style blowers" on the high-end Nvidia cards)[/quote<] I'm tired of this myth because it doesn't make sense: Temps [url<]http://www.techpowerup.com/reviews/ASUS/GeForce_GTX_780_Direct_Cu_II_OC/30.html[/url<] Noise [url<]http://www.techpowerup.com/reviews/ASUS/GeForce_GTX_780_Direct_Cu_II_OC/25.html[/url<] Imagine how quiter even the custom cooler would be if you made a fan profile that matched the stock cooler in terms of thermal performance...

        • HisDivineOrder
        • 6 years ago

        There are very few blowers out there that are superior to the Titan reference design. Open air custom coolers are one thing, but amongst blowers for the use cases where blowers make a lot of sense (and there are more than a few) there are few great alternatives to the nVidia premium blower.

        Certainly, AMD would be well served to have a good one designed. Perhaps they should use that IceQ or whatever design.

          • Arclight
          • 6 years ago

          It’s not too late for a R9 290XT edition with a modified version of the HD 7990 open air cooler (although it will probably never happen, they’ll be fine with custom coolers from board partners).

    • UnfriendlyFire
    • 6 years ago

    THG is trustworthy?

    Wat?

      • willmore
      • 6 years ago

      A decade ago they weren’t laughable, but now?

        • Airmantharp
        • 6 years ago

        It’s been long enough, that I think a decade ago they WERE laughable. They’re still laughable.

          • auxy
          • 6 years ago

          A lot of people say this these days and I don’t get it.

          What’s so bad about Tom’s?

            • Airmantharp
            • 6 years ago

            You’re young, hun.

            They were the epitome of half-assed reviews,stolen data, and bought and paid for conclusions. ‘Trust’ is something that they’ll not likely re-earn.

            • Bensam123
            • 6 years ago

            They used to botch reviews based on who was sponsoring them. I used to read Toms, but that was a decade ago before I switched to TR. Not sure what happens there anymore, but I’m pretty much TR and occasionally Anand now because of it. Most of what I want to read I can find here.

    • chuckula
    • 6 years ago

    A little more seriously… I highly doubt that AMD is intentionally trying to pull one over on the reviewers by going out of its way to cherry-pick cards for review.

    Instead, the variability is likely due to the fact that these cards are definitely running at the high-end of their thermal envelopes and that means even relatively small variations in the cooler performance or even ambient airflow can have larger-than-expected impacts on performance. Hopefully third-party coolers will mitigate these issues and AMD may want to investigate different cooling solutions for next-generation products to reduce these issues.

      • Ryhadar
      • 6 years ago

      I doubt it also.

      For a moment, let’s say AMD [i<]was[/i<] trying to pull a fast one. Look how easily and quickly a site discovered an issue. Either AMD's collective ego is bigger than reason or it's a quality control issue. Occam's razor and all that. Sucks for the people affected regardless though. I hope AMD and their partners make good for those people.

        • HisDivineOrder
        • 6 years ago

        I doubt AMD is trying to “pull one over” on people

        That said, don’t underestimate AMD’s “collective ego.” Do you really think they had no clue about the frame latency problem? I mean, no clue? No clue whatsoever? None at all? I think they knew something was amiss, but as long as it couldn’t be proven–hardocp’s “feeling” that it wasn’t smooth–they weren’t going to be bothered spending money to fix it. This is over the course of years of people saying, “Hey, nVidia feels smoother at lower fps’s.”

        A lot of things AMD is doing seems a little off. For example, releasing an absurdly power-sucking high performance CPU that is costing up in the realm of the E-line from Intel. They announce it as prebuilt only and make a lot of noise about it. A couple of months pass and they drop the thing by hundreds of dollars and suddenly it’s in every online store everywhere. “Good news, everyone!” the AMD professor starts.

        AMD did the first reveal on the R9 series in September, implying a product launch in early October. A preorder that was to come didn’t come and a few weeks turned easily into a month. Delay after delay, even affecting the R9 290. They can’t even get more than a fraction of the units one would want out to really call a launch hard.

        Last year, AMD released a CPU with an embargo on certain parts of the CPU’s review and certain parts were okay to review. They let sites release half-reviews that essentially discussed the great parts of the CPU and played them up in a great light. Then later, a month I think?, they let the sites release the rest of their reviews after the initial round had been out there based solely on the great parts… they let the rest of the CPU be reviewed in total. Naturally, the rest looked less favorable.

        AMD bought and paid for anandtech. Full stop.

        AMD is using Mantle as a way to shut out its competitors from the high performance PC gaming market. I doubt it’ll work, but they ARE doing it. This is far, far from the company that ATI used to be pushing open standards for the betterment of all PC gaming, including their competitors.

        I mean, honestly, there are a LOT of things AMD is doing that seem off compared to what they used to do. A lot of them stink. So do I think they intended to get one over on reviewers or users? No, but I wouldn’t be very surprised either.

        Why attribute to cleverness what could more easily be incompetence? I think they were desperate to get great reviews and they do not care what the long term consequences of this are. They don’t have much of a long term if they can’t get something going really quickly. So they’re throwing caution to the wind and taking a chance with high temperatures, high power usage, and pushing the limits on what is acceptable (in acoustics, in performance per watt, in per watts) to try and balance the scales. This is not that far away from the FX 9xxx series. (Good God, imagine putting the FX 9xxx series high end in the same system as tri-CF based on the R9 290X uber.)

        A pattern becomes clear. When the Radeon 7970 came out at the end of 2011/beginning of 2012, they had a marginal performance improvement over the Fermi 580 part of the time, but they had superior performance per watt. They marketed around that and they only conservatively clocked the chip. This later became a boon for a lot of users to argue that the 7970 and 7950 were overclocking cards that were in need of a little help to get up to their full potential. They charged high for these, but nVidia showed up later with a mid-range card swinging for the high end. AMD looked at the 680 and thought, “Why didn’t we just clock it high?” They later did with the GHZ card, but the damage was done.

        This time, they’re coming in late. So they pulled the same stunt with their 290 that nVidia did with the 680. They took a product that wasn’t quite at the level they wanted and they added a routine to make it regularly boost up to the level it needs to be in ideal circumstances. Ideal circumstances being scenarios that are easy to generate when you’re in an open air bench, like most reviewers. Hell, most of the time a lot of reviewers aren’t even going to run the card for more than a few minutes in a row before idling for a while, so even better.

        So they prioritized things that would benefit that, focusing on running as high a temp as possible since that would hardly show up any problems in an open bench or for a few minutes a go. Then they put in safety protocols to kick in as time went on in a more typical closed system that would slowly slide down from that performance level. Reviewers wouldn’t be likely to see it because they aren’t likely (at least at first) to use a system like that for hours at a time in a closed case.

        Is that “getting one over on people?” It seems like they’re just focusing on reviewers. Similar to how every Gaming Evolved win AMD has won this year has been the sequel or a followup from a company that made a benchmark of the last few years. They focused on almost exclusively the benchmarks that have been used over the last few years (Crysis 3, Total War Rome, Bioshock Infinite, etc) or games that would certain to become benchmarks. There’s nothing wrong with the strategy in that nVidia has done the same, but this was not an accident. It was a deliberate choice. Whoever is running Gaming Evolved certainly is no fool.

        But the hardware part seems partly based on a desire to review well and partly a desire to protect systems long term at the expense of performance. I just don’t think AMD expected the cards to hit the lows they are. I don’t expect AMD really thought much more beyond, “We must sell these cards and we must build some momentum for Mantle to have any impact. If we can’t get this going right now, we’re screwed.”

          • Pwnstar
          • 6 years ago

          Going for the essay-style reply, I see.

    • wierdo
    • 6 years ago

    It’s interesting to see a video card model that’s so dependent on cooling to perform, under ideal conditions that can be a great feature, but I don’t think the stock fan is gonna cut it.

    Would be curious how 290 and 290x would behave when temperature is kinda taken out of the equation with proper third party cooling, there seems to be a allot of headroom for overclocking from what little Anand tried with it.

    Would be more curious to see what people do with these chips in conjunction with some water cooling, maybe they can hit 1200mhz with some luck? I dunno.

      • Airmantharp
      • 6 years ago

      Those that have put them under water have reported crazy fast clocks, so there’s very little doubt that if a vendor figures out how to put away the ~350-400W of heat produced from the lower-performing parts with overclocking and corresponding voltage bumps, performance isn’t going to be an issue.

    • indeego
    • 6 years ago

    I kinda assumed this has been going on for decades now amongst lots of product categories. Kinda why Consumer Reports buys cars directly off the lot like any consumer would.

    • brucethemoose
    • 6 years ago

    AMD could’ve spent maybe $15 more on the cooler and boost performance while avoiding this indecent all-together. They poured hundreds of thousands into chip and board design, one would think they’d spend a little more time with the cooler for a massive GPU.

    • jimbo75
    • 6 years ago
      • Questar
      • 6 years ago

      Can we ban this clown yet?

        • NeelyCam
        • 6 years ago

        Yes We Can!

          • superjawes
          • 6 years ago

          The Hammer has fallen!

            • NeelyCam
            • 6 years ago

            [url<]http://www.youtube.com/watch?v=Z9BTZrFRYfw[/url<]

        • Ryu Connor
        • 6 years ago

        Done.

    • Firestarter
    • 6 years ago

    Sounds like they’ve been cherry-picking their review samples. That’s not unusual per se, but up until now all cards have always had a guaranteed base frequency, independent of how good or bad the particular GPU die was. Even the various boost schemes always had a guaranteed base. Now, with the performance so tightly coupled to thermals and by extension tightly coupled to individual die characteristics, this practice which was business as usual so far might come back to bite them in the behind.

      • Airmantharp
      • 6 years ago

      Well, their ‘base’ clockspeed is below 700MHz. That’s a big step down from the 900-1000MHz boost clocks that they MAY reach if the thermal headroom is there.

        • Firestarter
        • 6 years ago

        if they’d been advertised as such, the reviews would have been far more skeptical

          • Airmantharp
          • 6 years ago

          That’s off Anandtech’s review of the 290- and it was a pretty skeptical review!

        • Modivated1
        • 6 years ago

        So far this dramatic outcome has only been produced on one card, the claim is it has a problem. Time and reviews will tell if this is a common occurrence.

        We can’t accurately claim base clocks because they got one card to function that way. They had two and yet they could report the outcome for one, the other card is claimed to drop off like cards before it but there has been no info claiming that it had the same degree.

        Until either Tom’s Hardware is more forth coming and other reviews surface showing cards that were bought instead of supplied samples the jury is out. However, I won’t buy a card until this is resolved. I need to see a sound third party after market card with good noise, clocks, performance and decent heat control (it’s Winter time a little heat never hurt nobody 🙂 ).

    • anotherengineer
    • 6 years ago

    “Retail Radeon R9 290X cards may be slower than press samples”

    [url<]http://memecrunch.com/meme/FDZ9/what-you-talkin-bout-willis/image.png[/url<] Probably a VGA BIOS thing, since reviewers got an AMD ref card and you can't get those at retail anymore (excluding firepro's) So is it all cards or is it specific to XFX, Asus or others??

    • JohnC
    • 6 years ago

    I’ve seen that inconsistency. Easy to see when using MSI Afterburner – when playing some game like BF4, the GPU clock speeds on my EVGA Titan produce a smooth, straight line. When playing same game on my 290x – the GPU clock speed line has a sawtooth pattern. It does approach 1GHz, but only momentarily.

    Edit: I made a screenshot and separate forum thread, go check it out:
    [url<]https://techreport.com/forums/viewtopic.php?f=3&t=90345[/url<]

      • ChronoReverse
      • 6 years ago

      Interesting. Are you able to adjust the fan speeds? I wonder if you set the fan faster (louder) if it might fix it.

        • Airmantharp
        • 6 years ago

        Anandtech did a writeup exploring the fan speed effects on FPS- and they’re dramatic. Performance is coupled directly to cooling efficiency.

          • ChronoReverse
          • 6 years ago

          In that case, it’s looking like the crappy AMD stock cooler strikes again?

          If I were in the market for a 290(X) I’d wait for something like the Gigabyte tricool.

            • Airmantharp
            • 6 years ago

            Well, not really, that we know. The cooler is indeed very nice, but it’s nowhere near what’s needed for the thermal envelope AMD is pushing these cards to, and if they are indeed using significantly lower-binned parts then this is the expected result.

            It does remind you of the whole HD4830/HD5830/GTX465 fiasco though, doesn’t it?

        • JohnC
        • 6 years ago

        I haven’t tried that yet. In general I don’t like to screw up with stock speeds.

        • JohnC
        • 6 years ago

        Ok, I’ve manually raised fan speed limiter from stock 40% to 50%, it does get rid of throttling but (obviously) makes the card more noisy. Check out all relevant screenshots in my forum thread.

          • JustAnEngineer
          • 6 years ago

          Good job, John.

          What’s the corollary here? If you have to set the fan to the Uber speed to get the Uber performance, how much higher performance could you get with a really over-the-top cooling system on this same GPU?

            • Airmantharp
            • 6 years ago

            What’s with the implied criticism?

            Are you that disappointed that you have to make panning remarks at the one guy on this site that’s sharing his experience as he works with the card? I’d honestly taken you as a far more objective person.

            • JustAnEngineer
            • 6 years ago

            Only a crazed fanboy would think that “Good job” was a criticism.

            • Airmantharp
            • 6 years ago

            You don’t think that it could come off as sarcasm, when you ask ‘why are you implying the obvious?’ right afterward?

            • JohnC
            • 6 years ago

            I don’t know. There are plenty of crazy people with water cooling setups at Overclock.net forums, though, so if you are interested to know what this card is capable of with better cooling – try looking there 😉

          • Modivated1
          • 6 years ago

          Whew! Well that’s a relief for me, I was worried that there wouldn’t be a cooling solution capable of keeping this card at the posted speeds. Now I know it’s possible all we need now is a quieter solution.

            • Airmantharp
            • 6 years ago

            We’ve known that since the first reviews hit :).

            As Krogoth mentioned in the review thread, in a comment that’s currently at the very bottom yet highly up-voted, the power draw over the card’s contemporaries isn’t that much; we’ve had cards that draw more power that were easily stabilized and even significantly overclocked for the better part of a decade. The only concern is that the custom cooled versions are taking so long.

            Or maybe it just seems that they’re taking a long time, given that they’re needed now more than ever, yet we haven’t heard a peep :D.

            • JohnC
            • 6 years ago

            You can buy coolers from Arctic Cooling right now, both the “closed loop liquid cooler” and “multi-fan air cooler” option are available and compatible. It will cost you some more $$$, though, and installation might not be very easy.

          • Bensam123
          • 6 years ago

          I can’t help but feel this entire thread and responses is determined to push a certain line of reasoning. John is trying to push recirculating fans while his entire case is buttoned up tigther then a nuns-knickers (which of course doesn’t help a rear exhaust over a recirculating fan) and everyone else is trying overly hard to agree with it without giving any sort of criticism or suggestions besides Chrono.

          For instance, say, taking the side panel of his case off… if that helps remove the tape on all his intake vents which he decided to tape up. If you still have a problem, I’d say going for 43% based on your graphs.

            • Airmantharp
            • 6 years ago

            Unless his case is set up for negative airflow, taking the panel off or pulling the tape off of the cracks would only make things worse. Sadly, if he had to tape the cracks, then his case really is set up for negative airflow.

            Positive positive helps blowers and keeps the system clean- balanced airflow helps open-air coolers and requires a lot of care to get right. Negative pressure is just a bad idea :D.

            • Bensam123
            • 6 years ago

            Not everyone does things with positive or negative airflow in mind. He may have just been trying to keep hair and dust out of his case.

            • Airmantharp
            • 6 years ago

            And that’s what positive airflow with good filters does!

            • Bensam123
            • 6 years ago

            Or taping up all the intakes because it’s sucking dust and hair through them due to it being a negative airflow system.

            • Diplomacy42
            • 6 years ago

            negative pressure, positive airflow. airflow can’t be negative.

            even in a vacuum, the worst you can have is zero airflow.

            • Airmantharp
            • 6 years ago

            Finding quick ways to explain the amount and direction of airflow and relating differences in air pressure between the inside of an enclosure and the outside…

            ‘Negative’ in this scenario means lower air pressure inside the case than outside, positive the reverse. I have caught myself mixing up ‘pressure’ and ‘airflow’, though :).

            • Diplomacy42
            • 6 years ago

            Right, negative pressure means a pressure differential where the highest pressure is outside the case. Negative airflow, however, means nothing and “positive” airflow means that the fans are working.

            • JohnC
            • 6 years ago

            How about you stop assuming that YOU know my case’s airflow better than myself, huh? Or do your own tests for a change instead of posting your useless nonsense (like you do in almost every discussion)? I did the test with CASE OPENED, go look at the forum thread again:
            [url<]https://techreport.com/forums/viewtopic.php?f=3&t=90345[/url<] I even hid the case's panel in a closet and locked the closet's door so the evil side panel would not try to sneak around and sabotage any tests. Good enough for your highness?

            • Bensam123
            • 6 years ago

            Just because you own something or posses something doesn’t mean you know everything there is to about it dude. I mean for instance just because you’re in a body doesn’t make you a doctor.

            You did test it after I pointed it out. You hadn’t before I mentioned it. It was entirely possible the fan was starved for air and there was no way of knowing until you tested it. ‘Cause my case and I are awesome’ is not scientific testing.

            • Airmantharp
            • 6 years ago

            But he IS doing the testing, and responding to community comments and concerns.

            He doesn’t have to think that he’s awesome; he IS awesome :).

            • Bensam123
            • 6 years ago

            I thought we still worked on the premise of the scientific method here, not reputation? Hence all the testing Scott and Geoff do instead of just saying ‘this is awesome, this is why I think it’ and then writing a subjective piece.

      • cynan
      • 6 years ago

      I appreciate the corroboration of Tom’s Hardware’s findings. I for one put more stock in JohnC’s appraisal of the r9 290/x’s thermal issues than anything that could come off of that site.

      That said, I can’t say I’m all that surprised. We’ve known since Damage released his review of the 290x how close to the thermal limit AMD is pushing these things. Sure, Hawaii may fair a bit better at higher temps than its Tahiti predecessor, but 94 deg C, is well… just ask Paris Hilton.

      Furthermore, much as I hate to say it, there’s got to be a reason why AMD didn’t go with an advertised base clock on these things. It can’t just be coincidence.

      The only thing left to discern is whether aftermarket coolers from OEMs can at least guarantee the advertised 1 GHz clock rate for the majority of usage scenarios, at a decent decibel level. The way things stand with the stock blower… Well, let’s just say, maybe $399 for the 290 isn’t such the bargain basement price it first seemed, given the apparent necessity to have the cooling immediately replaced in order to get advertised performance with any sort of reasonable headroom.

      I’m sure the r290s are the bomb (and no, not just as in calorimeter) with custom water loops, but that doesn’t help the average customer very much..

    • Airmantharp
    • 6 years ago

    It’s THG, so huge grain of salt, but the results make sense given that these cards run at their thermal limits and throttle to stay there. The real question that I hope more reliable sources will be able to answer is if these ‘variations’ are truly the result of barely functional bins being welded to the cards, or if there’s some testing control that was overlooked.

    I’m much more inclined to believe that it’s the latter, but as others have stated, this needs some looking into.

    • Cyco-Dude
    • 6 years ago

    i [i<]couldn't[/i<] care less what thg posts. tr is a far more reliable source. there, now my level of caring can not possibly be any lower. -_-j

      • superjawes
      • 6 years ago

      The phrase is “I [i<]couldn't[/i<] care less."

        • cynan
        • 6 years ago

        “I could care less” was popularized about 50-60 years ago, mostly in the US, to mean that one does, in fact, care very little to nothing at all about a subject matter. The paradoxical nature of the phrase provides emphasis to how little one cares. It might not quite fit standard definitions of irony or sarcasm, but there it is. This comes up periodically in the comments section.

        Eg: [url=https://techreport.com/news/21804/windows-8-will-use-less-memory-than-win7?post=587605<]Here[/url<] and [url=https://techreport.com/blog/16849/laptop-hunters-more-like-desktop-replacement-hunters?post=401379<]Here[/url<] and even way back [url=https://techreport.com/news/11215/windows-vista-is-gold-release-set-for-jan-30?post=150501<]Here[/url<]. Phrases don't have to make sense at face value in order for them to be correct. Language evolves over time, and sometimes to include phrases or sayings that don't make a lot of sense, and in the rare extreme case, seem to convey the opposite of what they in fact do. You could just as easily argue that expressing that something is pretty great by calling it the "cat's pajamas" or the "bees knees" is nonsensical and stupid. And you'd probably be right. But they meant something to enough people at some point that we're stuck with 'em today. Such phrases are called an idioms. Some are more nonsensical than others. Deal with it.

          • torquer
          • 6 years ago

          Its amusing that you spent so much time replying only to sum it up with “deal with it.”

          Popularizing inaccuracy does not make it accurate. Words have meaning and the dilution of it to the lowest common denominator to fit in with current parlance is disturbing.

          It is properly written and stated, “I couldn’t care less.” Just because folks have decided to accept the incorrect version doesn’t make superjawes’ statement any less valid.

            • superjawes
            • 6 years ago

            I’ll defend some sayings if they morph into something that is still functional. For example: “all intents and purposes” versus “all intensive purposes.” I maintain that, while “intents” came first, the newer version is still accurate, for all intents and purposes 😉

            • auxy
            • 6 years ago

            They don’t mean the same thing and ‘intensive purposes’ makes you look ignorant and ill-educated.

            Likewise, “I couldn’t care less” is a hypercorrection. The phrase “I could care less” is facetious; as if saying “You/They should be grateful I am as concerned as I am, even as little as it is.” It is not wrong.

            • superjawes
            • 6 years ago

            “I could care less” is wrong because it does not specify the intent, which is to say “I don’t care” only with more sass. Theoretically it could mean that you care quite a bit, so if you [i<]really[/i<] care about something, technically speaking, you can care less 😉 As for "intensive purposes," the original phrase is a substitute for "essentially," essentially. "Intensive" is a strong adjective signifying importance, so "intensive purposes" is saying "for our important purposes." It might not be a 1:1 replacement, but it still serves the purpose the phrase is aiming for (unlike "I could care less").

            • cynan
            • 6 years ago

            Superjawes’ statement may be valid. But it’s incorrect. He is informing the OP that he (assuming “Dude” signifies male) has used a phrase incorrectly. This is not true. In the vernacular of many English speakers his usage of “I could care less” was correct usage.

            And it’s not about inaccuracy. It’s about the evolution of language overtime. The stance that something is incorrect because it doesn’t make sense at face value – when it in fact has, and continues to make perfect sense to enough of a critical mass of English speakers – is tantamount to illiteracy. Or at least ignorance.

            • superjawes
            • 6 years ago

            I think you misunderstand what I am arguing. I am saying that “I could care less” is a bad phrase, and even if it is being used in the correct setting, the incorrect wording undermines the purpose.

            “I couldn’t care less” would not only have been correct usage, but would better serve the purpose of the comment.

            • Cyco-Dude
            • 6 years ago

            ok, relax guys. my level of caring is now at zero. we can all go back to discussing the subject at hand: hot radeons.

            94c is pretty damn hot! it makes me think that they should have come with a gpu water cooler? like the ones they have for cpus (corsair, etc).

            • superjawes
            • 6 years ago

            Hot they are, but the temperature would not be as big of a deal if it did not undermine the performance.

            And you couldn’t make a liquid GPU cooler like that reference due to contraints. It works on CPUs because it is an aftermarket decision, and one that requires relocating the fans to a new location (which is the main draw). If you put something like that on a GPU, you would need to do the same thing or tie it into the CPUs radiator.

            • Cyco-Dude
            • 6 years ago

            i’m not so sure about that; didn’t some intel cpus (or amd cpus?) come with liquid coolers?

            moving on…i see some lamenting the poor performance of the reference cooler. business as usual, right? i’m sure we’ll see better cooling solutions from the various companies like we always have.

            • Airmantharp
            • 6 years ago

            They did, but they were just repackaged versions of other integrated water-cooler OEM’s products, and only came with CPUs that were already stratospherically over-priced.

            • cynan
            • 6 years ago

            Well, now you appear to be backtracking. You have every right to your opinion that the idiom “I could care less”, meaning “I don’t care a lot” is a silly/clumsy/confusing turn of phrase and better off not used. I may even agree. But that’s not what you said, or even remotely conveyed by simply stating “The phrase is “X””. This, to me, came off as a blatant invalidation of the OPs usage, which is not just unjust, but downright inaccurate, as it is also equally “Y”, whether you like it or not.

            You don’t get to make the rules. Nor do I. All we get are our opinions.

            Moreover, this isn’t a technical or scientific paper (where loose language and idioms in general would be inappropriate), these are comments on the internet. If you can’t be creative/colloquial//colorful with your phrasing here, where can you?

            To each their own. Variety the spice of life. To me, persecution for a bit of creativity, in what is essentially a form of creative writing (albeit simplistic) is just a bit presumptuous, small minded and at least as silly as the OP’s usage of “I could care less”.

            • Spunjji
            • 6 years ago

            The sad thing is I agree with you entirely; yet it still drives me nuts every time I hear someone say “I could care less”.

          • superjawes
          • 6 years ago

          Just because it’s popular does not make it correct. “I could care less” undermines the point of the phrase, so it ought to be forgotten.

          And mathematically, “I could care less” means:
          Care > 0

          While “I couldn’t care less” means:
          Care == 0

            • cynan
            • 6 years ago

            No one is arguing what the face value meaning of either phrase is. I agree with you on this 100%.

            Edit: I’m not even saying I like the phrase or agree with its usage. However, this usage now happens to be part of the English language, for better or for worse. Therefore, the original usage was correct.

          • JustAnEngineer
          • 6 years ago

          Stupidity with seniority is still stupid.

      • cynan
      • 6 years ago

      Caver! ;-D

        • Cyco-Dude
        • 6 years ago

        hahaha! it was edited after i read #6. i didn’t realize it was such a…HOT…topic. 😛

      • superjawes
      • 6 years ago

      [quote<]there, now my level of caring can not possibly be any lower. -_-j[/quote<] +1 for that!

      • Mr. Eco
      • 6 years ago

      Semantics issues aside, THG did provide useful information by reproducing repeatable throthling issue. It is just a valid info as the frame-pacing issues analyzed by other sites.
      Some two years ago THG hit the bottom; they improved a bit since then.

      • Arclight
      • 6 years ago

      They are the self proclaimed authority on PC hardware and we must fight the authority.

      [url<]http://www.youtube.com/watch?v=PKffm2uI4dk[/url<]

    • superjawes
    • 6 years ago

    Well this is dissapointing…please do follow up on this. That price/performance chart doesn’t mean anything if the retail versions don’t actually live up to their data points.

      • ChronoReverse
      • 6 years ago

      In particular, systematic testing is required.

      For example, if the retail cards are being run in a hot room while the test cards were in an air-conditioned room, obviously the thermal headroom would be different.

        • superjawes
        • 6 years ago

        That would also make a difference, as I believe TR still uses the open testbed chassis, and I imagine most users have an actual enclosure.

          • Airmantharp
          • 6 years ago

          At the same ambient temperature with a positive airflow enclosure, the cards should perform better. Testing blowers on an open bench does them no favors :).

    • chuckula
    • 6 years ago

    This is obviously a completely fabricated anti-AMD smear campaign.

    How do I know?

    THG claims to have been able to purchase the R9-290X at retail. What a crock.

      • Airmantharp
      • 6 years ago

      Could have stopped at ‘THG claims…’ 🙂

      • Meadows
      • 6 years ago

      Monty Python and THG.

Pin It on Pinterest

Share This