• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

ComputeMark - GPU Directx 11 Benchmark Stability Program

2410 at 1052/1312 5870
 
here's my score(Crossfired 5870 @ 900/1300):


attachment.php


attachment.php
 

Attachments

  • 5870Crosscompute.jpg
    5870Crosscompute.jpg
    58.4 KB · Views: 630
  • 5870Crosscomputea.png
    5870Crosscomputea.png
    191 KB · Views: 659
Last edited:
who ever it was earlier who said it worked on DX10 cards... no it doesn't. Doesn't even start. Plus I don't believe the scores are correct either.
 
the gtx480 obviously isn't very well optimised for directcompute 5.0 then. Nvidia obviously decided that gpu acceleration available to all with windows vista or windows 7 would not be used as much by devs as their own CUDA. Oh dears. That will be interesting when DC 5.0 starts to get used more. Let's hope OpenCL fares better for the gtx480, or fermi is stuffed.
 
here are my 2 5870's in crossfire

computmark.png
 
Looks like a bunch of crap to me. I seriously doubt that the 5970 does that much better than the 480.


The scores are perfectly on par with the theoretical values from each card.
This benchmark is just that, a proof-of-concept for theoretical performance.



The big difference is that, in any real-life computing software, the Fermi cards almost always achieve the peak theoretical values (they spent many hundred millions of transistors to get there), whereas ATI cards stop at around 30% of this value.
 
yeah that's why the results show this right...because Nvidia fanbois no more than everybody else...LOL
please continue with more reasons why Nvidia isn't doing so good..I need a good laugh
 
Here's my 5770's in Xfire --
4-18-2010 7-48-33 PM.png
 
yeah that's why the results show this right...because Nvidia fanbois no more than everybody else...LOL
please continue with more reasons why Nvidia isn't doing so good..I need a good laugh

well, AMD cards are better at this type of thing. I'd imagine really complicated tasks would probably suit fermi more. But in terms of what would be needed for gaming, i would say the AMD cards will have the upper hand, as this test shows. The test demonstrates dynamic smoke effects, something that could feasibly be added to a game.
 
Here is mine
attachment.php
 

Attachments

  • compmark 1.2 965-1200.JPG
    compmark 1.2 965-1200.JPG
    179 KB · Views: 630
@bobzilla
oh I think Nvidia will do just as well as ATI as soon as they start optimizing there drivers to be a little more compliant to such things..I just think it's funny cause Nvidia fans are in such disbelief that Low end cards competing against high end cards..At the moment. (i agree with you)
 
@bobzilla
oh I think Nvidia will do just as well as ATI as soon as they start optimizing there drivers to be a little more compliant to such things..I just think it's funny cause Nvidia fans are in such disbelief that Low end cards competing against high end cards..At the moment.

Yeah, i can't see a gtx480 losing to a hd5750 XD i can see it falling short of the hd5870 in stuff like this though, but only just. I read a review (can't remember the site) that was going over different types of calculations of both gpu's, and basically AMD won out on simple calculations overall by a small margin. But when the defecation hit the oscillation, the nvidia card pulled ahead significantly. Unfortunately for nvidia only the former will be used in games.
 
Back
Top