Memory Bandwidth on GTX285

Accelerated Computing CUDA CUDA Programming and Performance
1

Hello everyone,
I have been looking around to find the different mem speeds of the gtx285.
I came across this page:
[url=“GeForce GTX 285 Graphics Card Round-up | HotHardware”]http://hothardware.com/Articles/GeForce-GT...s-Card-Roundup/[/url]
all the cards mentioned there have mem bandwidths of ~165GB/s.

What is this Bandwidth between? System Mem and Device memory, or global to shared?

Could someone enlighten me.

Thanks

2

that is device-to-device memory bandwidth, computed by

(MHz Memory Clock Speed) x (512-bit Memory Interface) / (8 bits per byte ) x 2 (double data rate).

for example, ASUS ENGTX285

1.3 GHz x 512/8 x 2 = 166.4 GB/s

this number is just a theoretical maximum bandwidth, you should use SDK/bandwidthTest to measure effective bandwidth,

which is upper bound of memory-bound problem.

I don’t evaluate GTX285, but in my experience effective bandwidth is about 0.8 * theoretical maximum bandwidth