Tesla C1060 vs. GeForce 260 CUDA on a budget

Hey there,

I want to put together a linux test rig for doing CUDA and was wondering if the math capabilities of these two cards were the same? Not the speed or the amount of memory, but just the math. I’m interested in doing double precision as well as single precision floating point and as far as I can tell the kernels should output the same code if ran on either the a 260 or Tesla. I could be wrong but thats why I’m asking.

Also I was wondering if there was a comparison somewhere that covered the differences between these two cards. How much local memory for both and things like that. Thanks a lot.

They’re both GT200s, so their capabilities are the same. The differences are in the number of SMs, the amount of memory, and clockspeeds.

Appendix A of the CUDA 2.1 programming guide has a nice table listing the capabilities of most of the existing devices.