The Cause Of And Fix For Radeon R9 290X And 290 Inconsistency

Radeon R9 290 Gets The Same Treatment

If you saw our R9 290 coverage on Tuesday and decided that a $400 card delivering almost as much performance as the $550 Radeon R9 290X sounded tasty, then you were probably disappointed that we refrained from drawing a definitive conclusion about the press board’s representativeness. Let’s see if that trepidation was justified.

At launch, using Catalyst 13.11 Beta 8, Radeon R9 290 would spin up to about 2550 RPM under load, corresponding to a 47% PWM value. The same 47% on Sapphire’s R9 290 yielded just over 2300 RPM.

It’s no wonder, then, that Sapphire’s retail card bounced off of the board’s 662 MHz floor after just a few seconds of gameplay. By the end of a 100-second run, we recorded an average core clock rate of 726 MHz.

Manually scaling the card’s fan speed up to match our press board required a 50% PWM setting. But even at the same 2550 RPM, we were only seeing an average of 891 MHz compared to the press card’s 939 MHz.

Plotting out average performance shows us that the press-sampled R9 290 is 23% faster than our retail card. That’s exactly what I was afraid of, and the reason we held off on a judgment.

However, the new beta driver also imposes a new fan speed target on R9 290 as well. Instead of aiming for 40% PWM, like AMD originally planned, or 47%, which it later updated through software, it’s now aiming for 2650 RPM (more than 300 RPM faster than our retail card and about 100 RPM faster than our press board).

With Catalyst 13.11 Beta 9.2 installed, the press card maintains a fairly flat fan speed, while the retail board encounters strange spikes above 2700 RPM before tapering off and repeating itself. 

Clock rates certainly do come closer together, though. AMD's press board has zero trouble maintaining its 947 MHz peak clock rate. Meanwhile, the retail R9 290 from Sapphire averages 904 MHz. That's not quite as good, but it beats the hell out of the 726 MHz it was averaging right out of the box and the 891 MHz it managed after we manually goosed its fan to 50% PWM.

As frame rates go, the retail 290 pulls within 3.5% of our press board, which is the sort of variance AMD originally told us to expect.

Making Things Right

Am I at least satisfied that whatever oversight resulted in retail cards shipping out at varying degrees of lower performance is being actively addressed by the team at AMD? Yes; now gamers are getting the performance that got them so excited when these cards first launched.

Catalyst 13.11 Beta 9.2 makes Radeon R9 290X a little louder as it adds about 100 RPM to the "Quiet" mode we originally reviewed. The difference is noticeable, but I don't think gamers lacking headphones will be perturbed by the sound. But the R9 290 was already loud. So picking up an additional 100 RPM, even after AMD cranked the PWM from 40 to 47% prior to launch, is detrimental.

The bad news is that I really couldn't imagine buying an R9 290 equipped with AMD's reference cooler, particularly in light of today's update that adds even more fan speed and noise. The good news is that I have now have higher hopes for third-party 290s. With Catalyst 13.11 Beta 9.2, our Sapphire Radeon R9 290 is just as fast as Asus' Radeon R9 290X, tested on the previous page. If we could just get our hands on more aftermarket cooling solutions, I'm pretty sure we could chip away at the most compelling reasons not to buy these boards today.

This thread is closed for comments
330 comments
    Your comment
  • xiinc37
    This kind of thing is why I use msi afterburner. Just set your own fan curve.
  • tttttc
    I never understand why ANYONE will buy the reference design cards.... even for the GTX780, the aftermarket cooler is way better and quieter than the stock fan.
  • eldragon0
    Thank you so much for this article, I've been waiting on a write-up of this sort. There is one thing I'd love to see. Assuming the 290 and the 290x have the same layout, is it possible for you to strap that aftermarket heatsink onto each of them and give us a comparison of both of them at retail without them being horribly throttled? I'd love nothing more than to see a 290 and a 290x head to head at full speeds !
  • itzsnypah
    845427 said:
    This kind of thing is why I use msi afterburner. Just set your own fan curve.


    That is not how powertune works.

    It's: Get Hot -> Get loud -> Drop clocks -> Get as loud as necessary to keep GPU from melting.

    So having a 'custom' fan curve does nigh nothing.

    E: I wonder what would happen if you fed the card hot air. What happens when you're at 100% fan speed and still pushing 96c+? Does it shut down, clock down even more or melt?
  • FormatC
    Quote:
    is it possible for you to strap that aftermarket heatsink onto each of them


    It is impossible. The reason is simple: thermal clue. As I wrote in my article about the thermal grease: after the burn-in it is nearly impossible to remove the small heatsinks. The risk to destroy the card is too high. This aftermarket cooler is good and quiet but it is a real one-way ticket. You can't return ;)

    For addition - I've done the same thing with R9 290 cards and another benchmark before AMD has changed the driver. We worked hard to detect the reason for this big variances. But it seems that the difference between the R9 290 cards is a little bit smaller.

    This was before:
  • Quaddro
    Praise this card..
    Now we can see 780Ti and 780 with more reasonable price..;)

    Still waiting msi hawk version or directCu version of both of this card..hope that series will handle the heat..

    Spend $100 more just for cooler (mk-26 + 2 fans) on reference card, is not a really good option..
  • Quaddro
    -double post- :D
  • Nintendo Maniac 64
    I wonder if the rest of the performance difference now boils down to how the stock thermal paste was applied.
  • FormatC
    This I've tested too. No mentionable differences. May be my cards were assembled by a very accurate robot ;)

    The voltage is a little bit different:



    But this is not strange. Typical tolerance.
  • iam2thecrowe
    I cant stand all this variable clock rate BS ever since it was introduced in nvidia and amd cards. the card throttles in the most demanding scenario's, where you need it not to throttle, so the "boost clock" speeds are just there to inflate benchmark figures without any improvement in real world performance. On my cards, i force a stable clock speed with Nvidia inspector, negating boost clock, games run more stable and predictably, the way it should be.

    p.s. the add for nail fungus you have on your page toms, nearly made me vomit. please no more nail fungus adds!!!!!!!
  • eldragon0
    482859 said:
    Quote:
    is it possible for you to strap that aftermarket heatsink onto each of them
    It is impossible. The reason is simple: thermal clue. As I wrote in my article about the thermal grease: after the burn-in it is nearly impossible to remove the small heatsinks. The risk to destroy the card is too high. This aftermarket cooler is good and quiet but it is a real one-way ticket. You can't return ;) For addition - I've done the same thing with R9 290 cards and another benchmark before AMD has changed the driver. We worked hard to detect the reason for this big variances. But it seems that the difference between the R9 290 cards is a little bit smaller.



    Thanks for the reply! I'm still dying to see the the 290x does when it's fully unleashed. Keep up the good work :D.
  • s3anister
    Nothing wrong with the GPUs. Honestly I don't see any issue, if you buy the reference boards get a waterblock to go with it otherwise wait for the 3rd party solutions with better stock coolers.
  • natoco
    Sorry i did not catch that, could not hear you over your video card fan
  • Crashman
    1422415 said:
    I never understand why ANYONE will buy the reference design cards.... even for the GTX780, the aftermarket cooler is way better and quieter than the stock fan.
    Because to a CPU overclocker, the aftermarket cooler is crap? I just had the same style cooler on another card cause a factory-built PC overheat its CPU. As in thermal throttling. And the CPU was liquid-cooled!

    The type of cooler your praising blows heat into the case, rather than out of the case. They're just a bad design concept done right, as opposed to a good design concept that's poorly executed.
  • FormatC
    My wish: a R9 290/290X with a HIS DHE cooler (IceQ) like on 7950/70
  • Quaddro
    1422415 said:
    I never understand why ANYONE will buy the reference design cards.... even for the GTX780, the aftermarket cooler is way better and quieter than the stock fan.


    Well, reference card usually has the highest component quality..
    That means more durable and has longer life than common 3rd party cards (xfx, zotac, his, polor, and another lowly cheap brand)..

    i've still running my 4 years old 5850 reference card..overclocked to 1ghz since out of box, and cooled by mk-13..well, no issue until today, and still rock..
  • FunSurfer
    R9-290/290X + reference design cooler + Mantle (extreme card taxing) + august (mid summer) + no air conditioning = KABOOM!!!
  • scrumworks
    Storm in the water glass. Just because its AMD we are talking about here.
  • de5_Roy
    amd needs to be anal about quality check. take a page out of intel's q.c. playbook.
    the gpu is good, the ref. cooler is bad.
    to me, the press gpu still looks like a golden sample, the clockrate remains higher than retail ones throughout...

    will different cases affect hawaii performance with ref. cooler (or aftermarket cooler, when they launch)? imagine running a centurion cpu with reference hawaii boards in cfx.... (a scorpius gaming pc).
  • beavermml
    which means, whichever 3rd party that can develop a very good cooler will enable us to get more performance right?
  • jk47_99
    It has to be the TIM, but still something that should have been picked up by the QA process. It really is a poor quickfire response to just ramp up the fan speed and hope for the best. However in the interests of fairness, will you pick up a retail 780 Ti and make sure the performance matches the press card?
  • joditas
    Even with the new driver the 2 retail cards still aren't on par with press card. The differences are small but you can't ignore the impression here. Either their partners really drop the ball quality wise in manufacturing or there is problem with the chips, it's like Asus is 290X Lite, Sapphire is 290X Lite OC, and press card is regular 290X, the tiers still carry over to the new driver... I want to support you guys but you guys really need to step up to shake off the second-rate impression in many people's mind when come to quality control.

    THG what you need to do next is to test the power consumption of these cards with the new driver.
  • somebodyspecial
    388413 said:
    I cant stand all this variable clock rate BS ever since it was introduced in nvidia and amd cards. the card throttles in the most demanding scenario's, where you need it not to throttle, so the "boost clock" speeds are just there to inflate benchmark figures without any improvement in real world performance. On my cards, i force a stable clock speed with Nvidia inspector, negating boost clock, games run more stable and predictably, the way it should be. p.s. the add for nail fungus you have on your page toms, nearly made me vomit. please no more nail fungus adds!!!!!!!


    The difference is NV guarantees you won't go below what they advertise, and perhaps give you whatever is there extra free (there way ALWAYS means at very least you get what you paid for OR MORE). AMD seems to be the complete opposite of this strategy.
  • iam2thecrowe
    925801 said:
    388413 said:
    I cant stand all this variable clock rate BS ever since it was introduced in nvidia and amd cards. the card throttles in the most demanding scenario's, where you need it not to throttle, so the "boost clock" speeds are just there to inflate benchmark figures without any improvement in real world performance. On my cards, i force a stable clock speed with Nvidia inspector, negating boost clock, games run more stable and predictably, the way it should be. p.s. the add for nail fungus you have on your page toms, nearly made me vomit. please no more nail fungus adds!!!!!!!
    The difference is NV guarantees you won't go below what they advertise, and perhaps give you whatever is there extra free (there way ALWAYS means at very least you get what you paid for OR MORE). AMD seems to be the complete opposite of this strategy.


    good point, at least in reference to the 290(x) cards, nasty marketing, slows down when the going gets tough. performance is going to be extremely variable depending on your airflow setup/ambient temps. the 7xxx series and other 2xx series work the same as the nvidia cards with the boost clock. still BS though, the boost clock causes crysis games to crash, at least in the case of my gigabyte cards (although crysis 2/3 does warn on the disclaimer factory overclocked cards may not be stable in their games). Would be interesting to see long duration stability of these cards under load. 90c is too hot, they will develop micro cracks in the solder in under a year. i see many RMA's in the future for the 290 series.