I didn’t realize you were using NVLINK, as opposed to PCIe. The peer metrics are for PCIe connected GPUs. The NVLINK section is what will show the communication between GPUs in your system. The only thing to be aware of is that they are device-wide, so you want to make sure that you don’t have traffic generated by other applications running at the same time.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| In case of using peer memory, How can I measure the L1 or L2 cache's value on operating GPU? | 3 | 237 | March 18, 2025 | |
| cuda 4.0rc2 cudaMemcpyPeer(Async) performance issues | 11 | 13132 | May 3, 2011 | |
| Peer Device | 1 | 1095 | August 12, 2013 | |
| L2 read/write misses greater than requests | 11 | 3130 | May 11, 2011 | |
| problem with cudaMemcpyPeer() - won't do copying | 0 | 1466 | February 16, 2012 | |
| Peer-To-Peer Access with cudaPitchedPtr | 3 | 1130 | October 19, 2011 | |
| cudaDeviceCanAccessPeer returns false | 1 | 7348 | April 16, 2011 | |
| Peer-to-Peer access GPUDirect | 0 | 6146 | August 8, 2011 | |
| Partial fail of peer access in 8 Volta GPU instance (p3.16xlarge) on AWS -> huge slowdown | 32 | 3794 | March 10, 2018 | |
| identical code on multiple GPUs attached to the same board. how to do p2p memaccess? | 2 | 957 | June 12, 2013 |