[techreport]AMD issues statement on R9 290X speed variability, press samples

[Final8ty]

Based on feedback from the enthusiast community, we’ve implemented an all new PowerTune mechanism in the AMD Radeon R9 290 series that exploits the full capability of the individual GPUs rather than clamping performance to a least-common-denominator type of capability level. This has the advantage of improving overall performance but does result in some degree of performance variability. These changes will also result in some degree of run-to-run test variability based on environmental and operational conditions in un-controlled test environments.

The range of performance differential is not expected to meaningfully change the user experience but we’ve taken note of recent reports that the degree of variability is higher than expected. Reasonably we would expect the variability to occur both above and below the performance of the press samples, however it appears that most reported performances are biased towards the low side. We are actively investigating these reports and we will update when we have completed our investigation.

It is important to note that is it to be expected that the more thermally limited the setting the more variation can naturally occur. AMD Radeon R9 290X, by default, ships with two default BIOS settings for different acoustic levels and the "Uber" setting was designed to limit the level of thermal throttling and comparisons in this test mode are more consistent. Furthermore, AMD have implemented comprehensive PowerTune controls within the OverDrive panel of Catalyst Control Center and users can tweak to their own desired balance between acoustics and performance.

http://techreport.com/news/25733/amd-issues-statement-on-r9-290x-speed-variability-press-samples

 

[hawtdawg]

Why can't AMD just take a page from Nvidia and release a card without a god-awful reference cooler for once?

 

[thedosbox]

we’ve implemented an all new PowerTune mechanism in the AMD Radeon R9 290 series that exploits the full capability of the individual GPUs rather than clamping performance to a least-common-denominator type of capability level. This has the advantage of improving overall performance but does result in some degree of performance variability.

Huh, so a predictable level of performance is now a bad thing? Interesting bit of PR spin there.

 

[BallaTheFeared]

Reasonably we would expect the variability to occur both above and below the performance of the press samples, however it appears that most reported performances are biased towards the low side.

You don't say?

 

[24601]

Mine requires 64-70% fan to maintain stock boost levels @ 70-72 F ambient.

They obviously lied insanely hard.

 

[Slomo4shO]

The god awful reference cooler paired with a very hot chip will continue to bite AMD in the rear for some time even after non-reference models are released. Here is hoping they learn from their mistakes in the next generation (unlikely based on previous track record).

 

[Silverforce11]

Mine requires 64-70% fan to maintain stock boost levels @ 70-72 F ambient.

They obviously lied insanely hard.

Gaming?

I don't believe that for a second. Uber is only 55% and rarely throttles.

If its LTC mining, then yes, I experience the same, it requires 70% fan to maintain high speeds.

 

[thilanliyan]

And even at 60% it is loud. They really should have done a better job with this cooler.

 

[Slomo4shO]

Crossfire R290 (its bloody loud!)

I approve of your signature :thumbsup:


This seems to be a good alternative to a custom water loop:
http://www.youtube.com/watch?v=YattxxyWRh0

 

[Silverforce11]

And even at 60% it is loud. They really should have done a better job with this cooler.

Really, there's a massive thread in the review, we all agree the reference cooler is crap.

The only thing that is a beatup is dodgy reviewers doing Quiet Mode testing, and making a big deal out of variability when its limited to 40% fan speed.

 

[BallaTheFeared]

That and the magic golden review samples.

 

[Silverforce11]

I approve of your signature :thumbsup:


This seems to be a good alternative to a custom water loop:
http://www.youtube.com/watch?v=YattxxyWRh0

Yup, waiting for that. I have 3x Thermaltake 3 Extreme and Pro sitting around, they will be put to good use. Its very loud since im mining LTC 24/7 and in CF, the fans get up to 70% or more (its summer here) so yeah, its literally a jet take off.

But at 47% speed? Nope, hardly noisy at all. Once its in a case, 1 meter away, 47% is fine. Too many review sites love to put the dB meter next to the card on an open top.. hardly indicative of real usage.

 

[thedosbox]

Yup, waiting for that. I have 3x Thermaltake 3 Extreme and Pro sitting around, they will be put to good use. Its very loud since im mining LTC 24/7 and in CF, the fans get up to 70% or more (its summer here) so yeah, its literally a jet take off.

But at 47% speed? Nope, hardly noisy at all. Once its in a case, 1 meter away, 47% is fine. Too many review sites love to put the dB meter next to the card on an open top.. hardly indicative of real usage.

Point being, it's noisy relative to other cards measured the same way. And given the amount of airflow necessary to ensure they run smoothly at max...

Plus of course there's the different firmware between retail and review cards, but hey, that too must be the fault of dodgy reviewers.

 

[Gloomy]

Point being, it's noisy relative to other cards measured the same way. And given the amount of airflow necessary to ensure they run smoothly at max...

47% is the amount of airflow a stock 290 needs to run smoothly at max. That's why AMD bumped it up from 40% a little while ago, and why the 290 beats the 290X in quiet mode sometimes (lol)

 

[thedosbox]

47% is the amount of airflow a stock 290 needs to run smoothly at max. That's why AMD bumped it up from 40% a little while ago, and why the 290 beats the 290X in quiet mode sometimes (lol)

Indeed they did. And yet TR still found variability with the 13.11 beta 9v2 drivers.

http://techreport.com/review/25712/are-retail-radeon-r9-290x-cards-slower-than-press-samples/2

Our primary goal was to compare the performance of the retail and press sample 290X cards with the new Catalyst 13.11 beta 9v2 drivers, which attempt to equalize blower speeds. However, out of curiosity, we also decided to test our initial review unit and the HIS retail card with the older 13.11 beta 8 driver, which doesn't equalize fan speeds, to see how much of a difference that software change makes.

 

[RaulF]

I haven't had too much time with my 290. But I set the fan at 53 I think and now it holds the clock fine. At 47 I saw a bit of throttling but it was to like 925 or so not into the 800.

Now I am in beta drivers and bf4 is sort of a mess too so I am not going to point fingers yet.

And yes I find the card loud at 47%, but once I get into the game not even with headphones it kind of goes to the back of my mind.

 

[SiliconWars]

AMD must be laughing their heads off, all that wasted effort on FUD and they've ended up selling out on the 290's because of Litecoin anyway.


Stop the thread crapping and take the Litecoin talk elsewhere, NOW.


-Rvenger

 

[JDG1980]

AMD must be laughing their heads off, all that wasted effort on FUD and they've ended up selling out on the 290's because of Litecoin anyway.

Really worked out well for them, didn't it? A large portion of the reference cards are selling to people who don't care about noise because they're using them in makeshift mining servers. By the time that blows over, AIB custom designs should be available and anyone interested in gaming will just buy an Asus DirectCU or MSI Twin Frozr/Lightning, which won't have these problems.

I hope AMD makes nice Q4 profits on all the cards they're selling... we need a competitive marketplace, and that means AMD needs cash for R&D. Even if you decide to go with Intel and/or Nvidia products, you still benefit because competition helps keep them on their toes (as witnessed by the price drop of the GTX 780 immediately following Hawaii's release).

 

[Lonyo]

Huh, so a predictable level of performance is now a bad thing? Interesting bit of PR spin there.

Both AMD and nVidia (as well as Intel) have moved away from a "predictable level of performance" by using boost clocks.
It's the way things now are.

AMD do have a "predictable level of performance" in their lowest clockspeed which seems to be 662MHz or whatever things seem to bottom out at, but that's not a typical clockspeed in gaming.

 

[Pandora's Box]

Mine requires 64-70% fan to maintain stock boost levels @ 70-72 F ambient.

They obviously lied insanely hard.

I don't see a R9 290 series card in your sig...

Mine require about 60-65% fan speed to avoid throttling. But I'm running tri-fire and they are sandwiched together. Need to get a bigger case with more expansion slots so I can move one of the cards to the 4th PCI-E slot to give them breathing room. The cards are loud while gaming, yes. I don't hear them though, since I tend to play games with sound....

 

[ShintaiDK]

It cant be any news to AMD. When you make such a huge base/boost delta. Its because the chip variance is so big that 1Ghz is essentially a magic useless number for PR only, that the broad amount of cards cant do within the regulated specs.

 

[Lonyo]

It cant be any news to AMD. When you make such a huge base/boost delta. Its because the chip variance is so big that 1Ghz is essentially a magic useless number for PR only, that the broad amount of cards cant do within the regulated specs.

Chip variance isn't that big, most people agree it has a terrible cooler.
With a good cooler, it would be fine, it would run max speed, and also use less power than they currently do.

The chip is fine. The cooler is not.

 

[ShintaiDK]

Chip variance isn't that big, most people agree it has a terrible cooler.
With a good cooler, it would be fine, it would run max speed, and also use less power than they currently do.

The chip is fine. The cooler is not.

The cooler is the same on all cards, yet a huge variance. Its simply because the chips electrical properties got a huge variance as well. And AMD already showed how little faith they got in their product in this case with the base/boost delta. A baseclock may I add, that AMD do not want to list anywhere in their material.

So no, the chip is not fine.

 

[Skurge]

The cooler is the same on all cards, yet a huge variance. Its simply because the chips electrical properties got a huge variance as well. And AMD already showed how little faith they got in their product in this case with the base/boost delta. A baseclock may I add, that AMD do not want to list anywhere in their material.

So no, the chip is not fine.

I would not call 5% huge. Unless of course the inside of your case is 45c

 

[Lonyo]

The cooler is the same on all cards, yet a huge variance. Its simply because the chips electrical properties got a huge variance as well. And AMD already showed how little faith they got in their product in this case with the base/boost delta. A baseclock may I add, that AMD do not want to list anywhere in their material.

So no, the chip is not fine.

You know what else has variance? Coolers. Both the cooler itself (e.g. the base) as well as the TIM.
Combine a small natural variance in chips with a natural variance in coolers, and you get an fairly high variance.
Add in an individual's case variance (airflow, other components etc) and now you have 3 or 4 factors which combine to mess things up.

Slap on a good enough cooler that those variances no longer matter, and it's all fine.
If all the chips were perfect, then sure, the cooler might not matter, but if all the coolers were better, the chip variance wouldn't matter either. And it's quite easy to fix the cooler, as identified by the people who have put on a different cooler.