Quote Originally Posted by kl0012 View Post
I wonder why there are no official numbers on RV770/870 GPGPU perf in various realworld test (such as SGEMM, DGEMM). After all, AMD positioning GPU as application accelerator. As I understand AMD ACML supports matrix multiplication on GPU but the perf is pretty low (one says 300 GFLOPS on RV770). There are some highly specialized algorithms to extract more flops, but it is not clear if those algorithms are universal enough to use it in the real world.
If we take the 300 GFLOPS on RV770 on face value the RV870 maybe able to perform at-least 600 GFLOPS and thus the 2*RV870 "5970" can do almost 1200 GFLOPS.

When LRB comes out later, a RV970 maybe waiting and if AMD's current trend continues rv970 may as well have equal amounts of shaders as the 2*RV870 "5970" and thus may produce 1200 GFLOPS well at-least more than 1000 Gflops