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

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
http://www.hardwarecanucks.com/foru...roundup-asus-evga-gigabyte-galaxy-msi-21.html

I urge you to read this review. The Asus card should be beating the MSI and Gigabyte cards, but it isn't. Why is that? because of variation. That is not 1%.

*blinks*

You're looking at a review that compares factory overclocked aftermarket models to the reference model. OF COURSE they have higher clockspeeds, sometimes factory OC'ed models have up to 200mhz higher boost clocks than the reference. Of course factory OC'ed models perform better with higher boosts than reference.

Do note. None of those cards dip below the guaranteed base clock. In fact, none of them dip below the guaranteed boost clock. Conversely, the 290 and 290X do not have a guaranteed base clock. The boost clock on the 290/290X is also not guaranteed, it's an "up to" speed which it struggles to even hit at factory defaults.

Look closely at this chart:

GTX-660-ROUNDUP-82.jpg


I've told you this before. Kepler GPUs generally throttle 1-2 bins at most which is 13-26mhz. If you look at that chart, they are throttling 13-26mhz. 13-26MHz does not cause a meaningful difference in perceptible performance.

Meanwhile, here's 290X throttling:

bf4-retail-hot-clocks-645x491.jpg


You see a difference here? Kepler throttling by 13-26mhz over a course of 6 minutes. R9-290X throttling by 100-200 or more MHz in a period of 1 minute and 20 seconds. There's a pretty BIG difference. Kepler delivers consistent performance with 100% performance potential at factory defaults and auto fan. R9-290X just isn't delivering consistent performance at factory defaults due to excessive throttling, and apparently something was fundamentally different with press cards performing even higher than retail cards. I'm still at a loss to explain why that is, other than the press cards having a different BIOS.
 
Last edited:

KaRLiToS

Golden Member
Jul 30, 2010
1,918
11
81
Look closely at this chart:

GTX-660-ROUNDUP-82.jpg


I've told you this before. Kepler GPUs generally throttle 1-2 bins at most which is 13-26mhz. If you look at that chart, they are throttling 13-26mhz. 13-26MHz does not cause a meaningful difference in perceptible performance.

Meanwhile, here's 290X throttling:

Why are you comparing with Nvidia Kepler?

Oh yeah right, because Nvidia is so perfect, right? o_O

And by the way, this is from the LegitReviews review you linked first.

UPDATE 12/3/2013: Over the weekend we spent some time with the new Uber and Quiet vBIOS that AMD sent us for the Radeon R9 290X, but it appears to have lowered performance and we need to figure out what is going on with it before we do any temperature testing. For example in 3DMark Firestrike our performance went from around 9,700 to 9,500 with our open air test bench at 18C. Once again something is up and we need to find the time to figure it out. How many gamers are buying $500+ cards to run them on quiet mode in cases with poor airflow? The performance on the Radeon R9 290X starts to drop in extreme heat situations, especially when run in quiet mode, but who is this really impacting so far? It’s tough to continue to focus on this issue as it is really time consuming and tough to keep the ambient temperature correct between all the test runs.

They are not even sure what is happenning. They can't even keep the same ambient between each tests. The reviewers test sample is too small.

Just to let you know, my QuadFire 290x, if they are overclocked at 1200/1550, they will remain at 1200/1550 under 50'C no matter which BIOS is used.

I think everyone knows that AMD reference coolers suck. Its like beating a dead horse.

Anyway, why are you even concerned about this issue when you show you have a GTX 780? Is it directly affecting you? Why are you so mad when its not even affecting you?
 
Last edited:

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
Someone else brought it up, not me, so I was responding to potential misinformation regarding the performance over time differences between GK110 and Hawaii. Also, the throttling and performance over time comparisons between GK110 and 290X are relevant as they are directly competing products. Not to mention, consistent performance is something that consumers expect when they buy a PC product? Certainly nobody would go buy a 4770k and expect it to perform worse than what press reviews indicated to them.
 
Last edited:
Status
Not open for further replies.