Supercomputer cluster vendor Appro is a niche player in the cut-throat HPC market, and it's bringing the nVidia Tesla GPU and the CUDA programming environment to the knife fight with its new HyperPower clusters. The Tesla GPUs and their related CUDA tools went commercial last fall and they're been picked up by a number of …
800 watts? Scary?
You can get to that quite easily with a decently specced PC sporting a couple of high-end Nvidia GPUs in SLi.
You were expecting Nvidia chips to suddenly chew less power 'cos they were stuck on a maths copro board rather than a graphics card why exactly?
Flames, 'cos I'll bet it's like Satan's fireplace inside the box too.
From TFA "it's not surprising". What surprised me were the ratings on switches - e.g. the power consumption of a high end 48 port Catalyst is in the same ballpark.