The Radeon HD 5970: Completing AMD's Takeover of the High End GPU Market
by Ryan Smith on November 18, 2009 12:00 AM EST- Posted in
- GPUs
Crysis: Warhead
As Crysis is the only game that particularly struggled under even a single 5870, this is the only game we’ve gone ahead and captured data for at resolutions other than 2560.
The 5970 gains the distinction of being the only single-card setup we have tested that can break 30fps on Crysis at 2560 – it’s enough to at least push Crysis in to playable territory here, even though we’ll gladly take more performance if we can get it. So Crysis hasn’t quite fallen yet, but it’s very close. It’s also something where the 5970CF results would be quite interesting if we had them.
Interestingly, the 5970 manages to lose not only to the 5870CF, but the 5850CF by a very slight amount. The 5970 is clocked exactly the same as the 5850CF and has 2 more SIMDs, so we certainly aren’t expecting it to lose here. This may be driver related or it may be PCIe bandwidth related (each half of the 5970 only gets 8x bandwidth) but we aren’t sure which.
Meanwhile the GTX 295 sees the first of many falls here. It falls behind the 5970 by 30%-40%. The 5870 gave it a run for its money, so this is no surprise.
Finally, overclocking the 5970 to 5870 speeds nets you another 14%-18% in performance, which is quite interesting since the theoretical difference is only supposed to be 15%. But these numbers are repeatable; the 5970 likes being overclocked here.
Facebook
Twitter
RSS
114 Comments
View All Comments
tcube - Monday, March 1, 2010 - link
Well my thought is that if amd would release this card on SOI/HK-MG in 32/28nm MCM config at 4 GHz it would leave nvidia wondering what it did to deserve it. And I wonder ... why the heck not? These cards would be business-grade-able and with a decent silicon they could ask for enormous prices. Plus they could probably stick the entire thing in dual config on one single card (possibly within the 300 W pcie v2 limit)... that would be a 40-50 Tflops card and with the new GDDR5(5ghz +) it should qualify as a damn monster. I would also expect a ~2-3k$ per such beast but I think its worth it. 50Tf/card, 6cards/server... 3Pflop/cabinet... hrm... well...it wouldn't be fair to compair it to general purpose supercomputers... buuut you could deffinatelly ray trace render avatar directly into HD4x 3d in realtime and probably make it look even better in the process...srikar115 - Sunday, December 6, 2009 - link
i agree with this reveiw ,here a complete summary i found is also intrestinghttp://pcgamersera.com/2009/12/ati-radeon-5970-rev...">http://pcgamersera.com/2009/12/ati-radeon-5970-rev...
srikar115 - Tuesday, April 20, 2010 - link
http://pcgamersera.com/ati-radeon-5970-review-suma...xpclient - Friday, November 27, 2009 - link
What no test to check video performance/DXVA? DirectX 11/WDDM 1.1 introduced DXVA HD (Accelerated HD/Blu-Ray playback).cmdrdredd - Sunday, November 22, 2009 - link
Clearly nobody buying this card is going to put Crysis on "Gamer Quality" They'll put it on the max it can go. Why is AT still the only tech site in the whole world who is using "Gamer quality" with a card that has enough power to run a small town?AnnonymousCoward - Sunday, November 22, 2009 - link
Why does the 5970 get <= 5850 CF performance, when it has 3200 Stream Processors vs 2880?araczynski - Saturday, November 21, 2009 - link
i look forward to buying this, in a few years.JonnyDough - Friday, November 20, 2009 - link
why they didn't just call it the 5880.Paladin1211 - Friday, November 20, 2009 - link
Ryan,I'm a professional L4D player, and I know the Source engine gives out very high frame rate on today cards. The test become silly because there is no different at all from 60 to 300+ fps. So, it all comes down to min fps.
I suggest that you record a demo in map 5 Dead Air, with 4 Survivors defend with their back onto the limit line of the position of the first crashed plane. The main player for the record will be vomitted on by boomer, another throws pipe bomb near him, another throws molotov near him also. Full force of zombies (only 30), 2 hunters, 1 smoker, 1 tank attacking. (When a player become a tank, the boss he's controlling become a bot, and still attacking the survivors).
This is the heaviest practical scene in L4D, and it just makes sense for the benchmark. You dont really need 8 players to arrange the scene, I think using cheats is much easier.
I know it will take time to re-benchmark all of those cards for the new scene, but I think it wont be too much. Even if you cant do this, please reply me.
Thank you :)
SunSamurai - Friday, November 20, 2009 - link
You're not professional FPS player if you think there is no difference between 60 and 300fps.