chibi@1804:~$ cat \/etc\/os-release
\nNAME=”Ubuntu”
\nVERSION=”18.04.2 LTS (Bionic Beaver)”
\nID=ubuntu
\nID_LIKE=debian
\nPRETTY_NAME=”Ubuntu 18.04.2 LTS”
\nVERSION_ID=”18.04″
\nHOME_URL=”https:\/\/www.ubuntu.com\/”
\nSUPPORT_URL=”https:\/\/help.ubuntu.com\/”
\nBUG_REPORT_URL=”https:\/\/bugs.launchpad.net\/ubuntu\/”
\nPRIVACY_POLICY_URL=”https:\/\/www.ubuntu.com\/legal\/terms-and-policies\/privacy-policy”
\nVERSION_CODENAME=bionic
\nUBUNTU_CODENAME=bionic
\nchibi@1804:~$ nvcc -V
\nnvcc: NVIDIA (R) Cuda compiler driver
\nCopyright (c) 2005-2019 NVIDIA Corporation
\nBuilt on Fri_Feb__8_19:08:17_PST_2019
\nCuda compilation tools, release 10.1, V10.1.105
\nchibi@1804:~\/NVIDIA_CUDA-10.1_Samples\/1_Utilities\/p2pBandwidthLatencyTest$ .\/p2pBandwidthLatencyTest
\n[P2P (Peer-to-Peer) GPU Bandwidth Latency Test]
\nDevice: 0, GeForce RTX 2080 Ti, pciBusID: 42, pciDeviceID: 0, pciDomainID:0
\nDevice: 1, GeForce RTX 2080 Ti, pciBusID: 43, pciDeviceID: 0, pciDomainID:0
\nDevice=0 CAN Access Peer Device=1
\nDevice=1 CAN Access Peer Device=0<\/p>\n
***NOTE: In case a device doesn’t have P2P access to other one, it falls back to normal memcopy procedure.
\nSo you can see lesser Bandwidth (GB\/s) and unstable Latency (us) in those cases.<\/p>\n
P2P Connectivity Matrix CPU 0 1 CPU 0 1 NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled. 2990wx Ubuntu18.04.2 RTX 2080Ti x2 NVLink2slot CUDA 10.1 simpleP2P p2pBandwidthLatencyTest deviceQuery bandwidthTest<\/a><\/p>\n","protected":false},"excerpt":{"rendered":" chibi@1804:~$ cat \/etc\/os-release NAME=”Ubuntu” VERSION=”18.04.2 LTS (Bionic Beaver)” … \u7d9a\u304d\u3092\u8aad\u3080
\nD\\D 0 1
\n0 1 1
\n1 1 1
\nUnidirectional P2P=Disabled Bandwidth Matrix (GB\/s)
\nD\\D 0 1
\n0 528.52 4.18
\n1 4.17 527.72
\nUnidirectional P2P=Enabled Bandwidth (P2P Writes) Matrix (GB\/s)
\nD\\D 0 1
\n0 530.67 47.00<\/strong>
\n1 46.94<\/strong> 507.31
\nBidirectional P2P=Disabled Bandwidth Matrix (GB\/s)
\nD\\D 0 1
\n0 534.45 7.72
\n1 7.78 532.24
\nBidirectional P2P=Enabled Bandwidth Matrix (GB\/s)
\nD\\D 0 1
\n0 524.11 93.71<\/strong>
\n1 93.79<\/strong> 528.95
\nP2P=Disabled Latency Matrix (us)
\nGPU 0 1
\n0 1.31 16.16
\n1 21.55 1.66<\/p>\n
\n0 3.77 9.87
\n1 9.87 3.73
\nP2P=Enabled Latency (P2P Writes) Matrix (us)
\nGPU 0 1
\n0 1.32 1.59
\n1 1.87 1.66<\/p>\n
\n0 3.72 2.89
\n1 2.89 3.70<\/p>\n
\nchibi@1804:~\/NVIDIA_CUDA-10.1_Samples\/1_Utilities\/p2pBandwidthLatencyTest$<\/p>\n