Hot answers tagged


I believe this can be accomplished with the Intel Performance Counter Monitor (PCM) tools: As of v2.5 it supports PCIe monitoring on Xeon E5 processors. (ref) Looks like subsequent releases added support for more processor types: Haswell, Broadwell, Skylake.


So, I found out that apitrace can be partially used for similar tasks. At least, at this time this soft makes available to trace graphics API calls and show viewports content in D3D8 applications.


No, TenserFlow needs NVIDIA® GPU card with CUDA® Compute Capability 3.5 or higher. But your GPU only has a CUDA® Compute Capability of 2.1 GeForce GT 630M (see this list).


I don't think there is any. The best you can do is just copy the kernel and loop it for the block size you are running on CPU.

Only top voted, non community-wiki answers of a minimum length are eligible