I have a question to ask about some of the environment variables to get thread placement working properly on the phi with nested parallelism.
Recently, I want to port a complex program based on cpu to MIC. Because of the complex struct ,so I use the _Cilk_shared to manager the pointer to complex struct.
When I want to debug an OpenMP application on MIC I get the following warning :Can't load libomp_db library. OpenMP support is disabled.$threadlevel is set to "native".
I've been hunting down a performance problem in my native Phi MKL application and have discovered a surprising culprit.
Hello guys! This is my first post here so apologies so far.