[dead]
The mean means nothing: data visualization to debug a latency problem
21–24 of 24 posts
Re: The mean means nothing: data visualization to debug a latency problem
#22The author is spot on. As an HPC performance engineer and computational artist, I could tell numerous stories about it, but I'll summarize three: The common method for evaluating MPI performance on a supercomputer is with the OSU microbenchmarks, and with the -z option will print tail latencies. These are far more useful values to predict real-world performance on the big machines. Second, while means DO say somethin…
If you don't mind me asking, how does one get into HPC performance these days?
Re: The mean means nothing: data visualization to debug a latency problem
#23The author is spot on. As an HPC performance engineer and computational artist, I could tell numerous stories about it, but I'll summarize three: The common method for evaluating MPI performance on a supercomputer is with the OSU microbenchmarks, and with the -z option will print tail latencies. These are far more useful values to predict real-world performance on the big machines. Second, while means DO say somethin…
If you don't mind me asking, how does one get into HPC performance these days?
The nice thing is that every gpu is like a little distributed-memory machine, and programming it for optimal memory locality (registers, shared, global) and bandwidth is an essential skill. One step up is using multi-node distributed memory machines, which you could set up at home with openmpi, but is often easier to get access in an academic setting. In short, learn how to use CUDA and MPI. That's my personal take on "HPC", though obviously it can mean different things to different people.
Re: The mean means nothing: data visualization to debug a latency problem
#24Earlier quoted context omitted.
If you don't mind me asking, how does one get into HPC performance these days?
The nice thing is that every gpu is like a little distributed-memory machine, and programming it for optimal memory locality (registers, shared, global) and bandwidth is an essential skill. One step up is using multi-node distributed memory machines, which you could set up at home with openmpi, but is often easier to get access in an academic setting. In short, learn how to use CUDA and MPI. That's my personal take o…
appreciate the perspective, thanks!