But how do you sleep at night knowing there are people out there with better hardware than you
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
NVIDIA GTX 480 makes benchmarking debut, matches ATI HD 5870 performance (video)
- Thread starter richb93
- Start date
But how do you sleep at night knowing there are people out there with better hardware than you
And with no AA and AF which must have been done for a reason to show Fermi in its best light.
That (should be) rubbish, it might have some impact on edge filtering through deferred shading but heaven doesn't use that. The method of implementing tessellation should have no impact on AA and AF performance unless for some bizarre reason tessellation and AF suck up so much memory bandwidth between them it has an impact but I can't see that happening with the architecture.
Maybe your right Rroff but I can't help but think why they didn't do the test with some AA & AF because if what you say is true it doesn't make sense because no one with a top of the range gpu doesn't use AA... (Unless your playing BM with an ati card)
Hard to pin down a reason tho it does raise questions... I would say it would be one of these:
Trying to show purely tessellation performance (but then why use high shaders - unless tessellation only enables with high shaders not something I've tested)
That particular card doesn't have fully functional hardware ("lot" of A2 parts worked but with bugs/problems)
They are still working on AA/AF implementation and performance isn't representative of final (my bet)
The benchmarker is an idiot
Also from my experience instrumented debug mode drivers can be upto 40% slower than the non-debug versions - which is why NDAs exist to prevent people releasing results that aren't representative of the final performance/quality.
Trying to show purely tessellation performance (but then why use high shaders - unless tessellation only enables with high shaders not something I've tested)
That particular card doesn't have fully functional hardware ("lot" of A2 parts worked but with bugs/problems)
They are still working on AA/AF implementation and performance isn't representative of final (my bet)
The benchmarker is an idiot
Also from my experience instrumented debug mode drivers can be upto 40% slower than the non-debug versions - which is why NDAs exist to prevent people releasing results that aren't representative of the final performance/quality.
Last edited:
The bottom line is that if they had something to show off they would....they really don`t - even given all this extra time.......if it they could run Crysis with full candy @ 80fps they`d be dangling it like a big willy....
.....after very careful analysis I`ve come to the very important conclusion.....
Nvidia is a poof
.....after very careful analysis I`ve come to the very important conclusion.....
Nvidia is a poof
But how do you sleep at night knowing there are people out there with better hardware than you
im still running AMD dual core.. still on 8800gt
all games run on high/medium why upgrade ? 
Associate
- Joined
- 18 Oct 2005
- Posts
- 661
I hope the card doesn't suck and doesnt cost more than £500
If fermi comes out and reduces the prices of the ATI cards, I'm happy as anything.
I can honestly say I can't see nVidia topping ATI, even with fermi. As others have said why would they show a benchmark WITHOUT any eyecandy-goodness on their top-end card?
I've always liked and preferred the green team but bought whoever I perceived had the best products out within my price range at the time. New upgrade soon = 5850
I can honestly say I can't see nVidia topping ATI, even with fermi. As others have said why would they show a benchmark WITHOUT any eyecandy-goodness on their top-end card?
I've always liked and preferred the green team but bought whoever I perceived had the best products out within my price range at the time. New upgrade soon = 5850
That won't be it. It does however make sense if the raster operators aren't up to scratch, however, as as far as I can remember, MSAA works per each polygon edge rather than for every pixel or for silhouette edges (like in edge detect AA) so with higher resolution models (i.e. those generated through tessellation), the workload increases. That would (possibly along with Eyefinity) possibly explain why ATi have doubled the number of raster operators on their 5 series cards, at least.