Test On GTX280

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi everyone.
Actually I am working on a neural network simulation (it’s LISSOM, a SOM variant).

[url=“http://homepages.inf.ed.ac.uk/jbednar/rflissom_small.html”]http://homepages.inf.ed.ac.uk/jbednar/rflissom_small.html[/url]

So far, I’ve been able to test it on my:
8600M GT
8800GTS (320mb)

And on others’ GPUs:
Quadro FX 5600 / Tesla D870

Now I’d need to test it on the new GTX280, as the performance speedup it may provide would be very useful.
I cannot access a 280, thought; would someone with a GTX280 be so kind to help me doing two short tests?

Thank you in advance.

PS: It would also be nice to test the simulation on a Tesla S870, to study how it scales on multiGPU systems, but I think it would be much harder to find it. :P

-Giacomo Spigler

2

I’ll be glad to help out. We have achieved speedups of 1.7 to 2.0 consistently
for GPU kernels (finite-element calculation and Voronoi mesh generation).

We do not have a Pci-express 2.0, so if that is an issue, I am sure somebody else
will reply. Thanks.

Gordon

3

PCI speed isn’t an issue, as all calculation is done onboard (inputs to the network are generated directly on the GPU, too).
I’ll contact you by PM.

4

It would be very interesting if you share with us the results :) .