I have but i think its in rare cases TBH.
Printable View
AMD (64 Viewing)
CPU's & Motherboards
Intel (113 Viewing)
CPU's & Motherboards
ATI (120 Viewing)
GPUs, Graphics Cards, and Drivers
Nvidia (3 Viewing)
GPUs, Graphics Cards, and Drivers
:rofl:,am not the only one looking forward to new series :)
Thanks a bunch 'postumus' for the benches. I'm so going to get this card.
Any word on the HD5770 performance?
Crysis
Run #1- DX10 1920x1080 AA=16xQ, 64 bit test, Quality: VeryHigh ~~ Overall Average FPS: 29.89
Just chews through the AA ;)
Anyway it's back to work tmrw so I think my lil preview is over, thanks again guys for the helpful links, my small hd is now full of demo and benching material ;)
I got
26.6fps no AA
21.62fps 4xAA
1080p Very High
with GTX285 SSC + i7 @ 3.6ghz
nvidia 190 drivers Vista Ultimate x64
Postumus, how did your framerate increase with 16xQ? O_O
Also how did you set this in the game itself or the ATI Control Panel? The benchmark is based on the game values so you have to restart the game to set the settings. Forcing values at driver level seem to have zero effect (16QAA supersampling/16xAF did not change my framerate at all). I ask this because you have an AF setting and I don't see the game has any setting to change AF, only AA...
Since 16Q is not a native ATI AA but rather Nvidia it could be a bug and simply defaulting to the next highest (8x).
so whats the verdict
it's just a touch faster than 4870X2? correct
man that thing will just wipe the floor with nvidia in quadfire in vantage looking at those numbers so far
Again as I asked earlier in the thread, how? I run 2560x1600 2x AA at around 35-40fps during races maxed in-game options 16x AF, sometimes up to 45 if not many cars on screen, in the full NFS Shift game. My system is a GTX 280 1GB, i7 920 HT on @ 3.72ghz, 6GB DDR3, X-Fi XtremeGamer, Asus P6T X58. Something smells funny with your statement.
Can anyone tell if this image is a fake? It's supposedly a RV870 die shot, but the interesting thing is that people (tweakers, rage3D) are counting 2560SPs. It would make a lot of sense, given the rumors of TMSC 40nm problems, that even with the larger die ATI could get a great yield if they only enabled 62.5% of the compute and texturing resources on RV870. After yields improved they create a full on 5890 SKU to go against G300. On the other hand, we've already had a fake die shot coming out of chiphell.
That's a fake....
The real die shot was shown in the slides on ATI's strategy...
Rumour says the 5900 will have 2560 sp's
Nvidia camoflage!
http://www.techfuels.com/attachments...assified-1.jpg
I know it's not the EXACT same engine, but throwing this out there - http://www.xbitlabs.com/articles/vid...0_8.html#sect2
That doesn't seem too far off :shrug:. Looks like ~30-35% boost over a 4890 AND it's using beta drivers.
Now THAT's amazing :clap:. I don't think any other card could ever run 16xQ AA in Crysis and get playable framerates, never mind at 1920x1080 :eek:. Fantastic results.
Thanks for all the posting and testing you've done, I appreciate it. :up:
yes its fake
http://forum.beyond3d.com/showpost.p...postcount=3035
i just woke up -,-
looks like that game is optimised better for ATI cards for superhigh res