via HPC Anwers
Chris at HPC Answers has an interesting post addressing the question “Are GPUs the next wave in HPC?“
As much as I’d prefer to see someone like ClearSpeed succeed over GPU-based general-purpose computing, I’ve seen enough in this industry to understand commoditization, volume, and market penetration. I believe a more likely scenario is that CPU + GPU will indeed become standard in blade-based clusters aimed at technical computing applications.
I agree, on both counts, unless the FPGA guys can move us away from having to code algorithms in hardware description languages and performance is a lot better (because CPU + GPU will probably be cheaper). Advances like SystemC may address the first point; the jury is still out on the second.
Given their energy advantages (10W “average” for the ClearSpeed co-processor versus over 100-200W for some high end GPUs), I wouldn’t discount ClearSpeed just yet, though.