fanlu commented on issue #8861: The first GPU cost more memory than the others in version 0.12.1 URL: https://github.com/apache/incubator-mxnet/issues/8861#issuecomment-352621939 @eric-haibin-lin When I use one gpu and batch size 30,the nvidia-smi log is +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 0 48806 C python 20365MiB | +-----------------------------------------------------------------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 1 50097 C python 19885MiB | +-----------------------------------------------------------------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 2 50914 C python 20007MiB | +-----------------------------------------------------------------------------+ +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 3 51716 C python 19855MiB | +-----------------------------------------------------------------------------+ But when I use 2 gpus and batch size 60,the problem also exist +-----------------------------------------------------------------------------+ | Processes: GPU Memory | | GPU PID Type Process name Usage | |=============================================================================| | 2 52525 C python 19989MiB | | 3 52525 C python 9383MiB | +-----------------------------------------------------------------------------+
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services