Hi, I recently tried to program with pycuda + CUDA8.0 + mpi4py,and I found that my desktop (with 1 i7 4770 and 1 GTX 1080)computing faster than the school’s HPC platform(with 2 Xeon E5-2660 v3 and 2 Tesla K80),We think this is because the performance of a single CPU core of E5-2660 v3 is not as good as i7 4770 and one GPU defaults to socket one CPU core. So,the question is how can I get an mpi process to call a GPU and multiple CPU cores?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Using GPUs on high performance machines | 4 | 1278 | February 8, 2013 | |
| Assigning GPUs to specific CPU cores using MPI and Server 2008 | 0 | 1901 | June 30, 2010 | |
| Question about CUDA+MPI | 3 | 2772 | March 13, 2018 | |
| maximum number of GPU in quad core CPU | 0 | 5452 | June 8, 2008 | |
| Multi-CPU + 2 GPU on computational expensive process | 0 | 428 | June 28, 2021 | |
| using all 4 GPUs in S1070 from multi-core cpu? how | 11 | 32663 | December 13, 2010 | |
| Example on MPI + CUDA on Two CPU and Two GPU node | 1 | 4077 | September 7, 2011 | |
| More cores than GPUs | 4 | 3710 | June 2, 2009 | |
| Programming on Multiple GPU and Multiple CPUs | 0 | 728 | September 5, 2011 | |
| MPI + CUDA Problem | 0 | 6522 | October 3, 2011 |