kaixiangjin
Results
2
comments of
kaixiangjin
> > it seems like the model inference images one by one, not as a whole to inference. > > Parallel processing is only performed when there is still surplus...
> GPU resources contains lots of things: register, l1, l2, memory bandwidth, shm, cuda core/tensor core etc. Usually need do experiments. > > It can be roughly viewed through nvidia-smi...