8800GTX Shadermark results
- Thread starter Geeforcer
- Start date
- Status
Anyone know what gpuinstr is? He's posted full results and I can guess what some of the numbers mean, but they're not really meaningfull without a comparison point.
http://www.hardforum.com/showpost.php?p=1030169416&postcount=417
http://www.hardforum.com/showpost.php?p=1030169416&postcount=417
Sounds like billion shader instructions executed per second.
The Fillrate Tester results show that PP runs at the idential speed to normal, and the shadermark results say that both PP and normal have full FP32 (s23e8) precision. So no, no difference (and it doesn't seem you can even force it to lower precision).
Yeah, the drivers that shipped with the cards must be over a month old at this point.
Mintmaster
Veteran
Man, this sucker is a BEAST.
wtf are you going to do with 70GPix/s z-fillrate?
You could have a 10x true overdraw (without culling) on a 10K x 10K uberscreen and you'd still be realtime! Not that I'm complaining...
Arithmetic and texturing perf looks to be over 2 times the G71 (128 scalar MAD+MUL units at 1350 MHz would explain it), but I'm not seeing enough differences in shading characteristics from G71 that would suggest a very different base architecture. Texture decoupling, if it did happen, doesn't seem to have helped one shadermark test over the other (relatively speaking), at least not to the degree of R420->R520. For similar reasons I'm doubtful about truly independent scalar ALUs.
Not that any of this really matters, as it's insanely fast. The VS also looks speedy for a non-unified design. IMO, R600 will have a tough time matching this unless ATI also made GHz shader engines. Hopefully AMD can help them out in this area for next-next gen.
I might make the switch to green this winter!
wtf are you going to do with 70GPix/s z-fillrate?
You could have a 10x true overdraw (without culling) on a 10K x 10K uberscreen and you'd still be realtime! Not that I'm complaining...Arithmetic and texturing perf looks to be over 2 times the G71 (128 scalar MAD+MUL units at 1350 MHz would explain it), but I'm not seeing enough differences in shading characteristics from G71 that would suggest a very different base architecture. Texture decoupling, if it did happen, doesn't seem to have helped one shadermark test over the other (relatively speaking), at least not to the degree of R420->R520. For similar reasons I'm doubtful about truly independent scalar ALUs.
Not that any of this really matters, as it's insanely fast. The VS also looks speedy for a non-unified design. IMO, R600 will have a tough time matching this unless ATI also made GHz shader engines. Hopefully AMD can help them out in this area for next-next gen.
I might make the switch to green this winter!
Right now, I am frantically looking for either Shadermark G71 results OR conclusive evidence indication what, if any, effect does resolution have on the results in order to directly compare the G70 and G80. The architecture does seem to give a very good account of itself VS R580. BTW, I think the preponderance of evidence at this point suggests unified design.
Meaningless without context. There are a number of applications that are CPU-bound for a particular CPU-GPU-resolution-quality combination, but, by itself, there is no such thing as CPU-bound GPU.
(Pendantic? Maybe, but these blanket statements are annoying and have no higher signal/noise ratio than the articles of you-know-who.)
Pedantic? Most certainly. I would certainly think its possible for you to at least imagine the GPU in its typical job (playing games!) and there you would certainly need a CPU that could keep up with it.
I just upgraded from a 19" to a 24" screen. Suddenly, all my games become GPU limited. Not surprising of course, but the best way to show that a CPU-bounded GPU doesn't exist.
I wonder how many common, popular games out there are CPU bounded at 19x12 on a 8800GTX. Probably not many.
What is your CPU? How do you know, unless of course you've done a wide range of testing, that you would not gain from the extra CPU power?
So you think someone cannot see what the bottleneck in only one system is with while running some tests? You don't need to compare a gazillion systems to see if a system is CPU bound; one indication would be that you all of the sudden get in resolution X GPU specific functions nearly for free and the next best riddle is what a four legged animal walking on a hot tin roof might be
Yup. You can gain additional performance at high resolutions with a faster CPU. But the effect is minimized compared to lower resolutions when the application is purely CPU bound.
Chris
- Status