• 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.

DirectCompute benchmark

Lightnix said:
No idea, but 'something' is clearly up, look at the scores posted not in comparison to the geforce cards, but in comparison to the radeon 5700. They're no higher. I suspect the benchmark is naively written and doesn't really take advantage of vectorisation properly, hence the low scores and lack of performance scaling on the Radeon cards. That, or something is really, really wrong with AMD's directcompute implementation. I don't think that's the case though, as according to Anandtech, the 5800 series run the Nvidia wave demo much faster than the GTX200 series... I suspect it's just a bit of a crap benchmark, in all honesty.

http://www.anandtech.com/video/showdoc.aspx?i=3643&p=8
Click to expand...

The wave demo was pretty crap too - didn't even use multi GPUs on the nvidia side and is deffinatly not any good for performance comparisions... whether this is or not is another story... which we are yet to find out.
 
its the same as comparing these results in heaven benchmark, one at dx10 and one at dx11, your not even going to get anywhere near the same results

dx10.jpg
dx11.jpg
 
No it isn't - do you understand what directcompute, opencl, GPGPU is about?

Sure DX11 directcompute and DX10 directcompute have different limitations and features... but according to the blurb that comes with the test this is just running the same sample FFTs and some bandwidth tests using whatever directcompute API is available.
 
People need to get the head's around API's and API features. The bench requires DX11 because it's a DirectCompute benchmark, which isn't supported under DX10 or lower.

As far as I know the scores are comparable, meaning they are doing the same work load, but it appears this benchmark has been written in a way that favours Nvidia hardware a lot.
 
well stick with your gtx260 then Rroff, and enjoy higher numbers while everyone else enjoys dx11;)

and no i dont understand directcompute, opencl, GPGPU is about and never claimed i did :D
 
I plan to stick with the 260 til theres a need for DX11 and then I will upgrade to whatever the best card is for it at the time.
 
Going forward, one specific issue for DirectCompute development will be that there are three levels of DirectCompute, derived from DX10 (4.0), DX10.1 (4.1), and DX11 (5.0) hardware. The higher the version the more advanced the features, with DirectCompute 5.0 in particular being a big jump as it’s the first hardware generation designed with DirectCompute in mind. Among other notable differences, it’s the first version to offer double precision floating point support and atomic operations
Click to expand...
http://www.anandtech.com/video/showdoc.aspx?i=3643&p=8
 
deffinatly - they have the potential for some very high single precision numbers in theory with the architecture.
 
Also some 2x 295GTX or 3x 285GTX scores would be great :D as well as 5870 crossfire - assuming its multi GPU scaling.
 
ahh its looks like it dosnt take into account the gpu is throttled down to 157 mhz in destop, is there a way to stop the throttling ?
 
hyperst said:
ahh its looks like it dosnt take into account the gpu is throttled down to 157 mhz in destop
Click to expand...

The CCC was left there meerly to show my clocks, ill run it again and check to see if it adjusts the clocks when the benchmark is running.

EDIT:

The clocks to adjust when the benchmark is running.
 
Also one in the eye for those who say HT is never good for more than a 10% boost - those i7s are doubling my CPU score clock for clock.
 
You must log in or register to reply here.
Back
Top Bottom