Xuechao Wei

weird pin thread to cpu performance

5

Hi @clamchowder, I want to pin thread to cpu when measuring bandwidth, but I found that there seems no such facility under the non-numa mode. So I just borrow this...

Hi @clamchowder, I found in MemoryBandwidth.c, under NUMA mode, there is a hardcode number (https://github.com/clamchowder/Microbenchmarks/blob/master/MemoryBandwidth/MemoryBandwidth.c#L620): As the comment indicates, it may be for KNL. So should we recover the `numaNodeCount`...

Hi @abhibambhaniya, I found one suspicious bug in function `get_offload_system`, in file `GenZ/LLM_inference/utils.py`. The unit of `total_memory_req` passed to this function seems to be `MB`, while `total_device_memory` is clearly `GB`...