Using GPUs with Slurm: Difference between revisions

Using GPUs with Slurm (view source)

Revision as of 17:52, 3 August 2023

83 bytes added , 1 year ago

Models with GB

Plstonge

cc_staff

782

edits

@@ Line 37: / Line 37: @@
 !                              CPU cores !! CPU memory !! GPUs
 |-
-| Béluga            || 172 ||  v100 ||  40 || 191000M ||  4 || V100-SXM2 || 70 || 16 || All GPUs associated with the same CPU socket, connected via NVLink
+| Béluga            || 172 ||  v100 ||  40 || 191000M ||  4 || V100-16gb || 70 || 16 || All GPUs associated with the same CPU socket, connected via NVLink and SXM2
 |-
-| rowspan=3|Cedar   || 114 ||  p100 ||  24 || 128000M ||  4 || P100-PCIE || 60 || 12 || Two GPUs per CPU socket
+| rowspan=3|Cedar   || 114 ||  p100 ||  24 || 128000M ||  4 || P100-12gb || 60 || 12 || Two GPUs per CPU socket, connected via PCIe
 |-
-|                      32  || p100l ||  24 || 257000M ||  4 || P100-PCIE || 60 || 16 || All GPUs associated with the same CPU socket
+|                      32  || p100l ||  24 || 257000M ||  4 || P100-16gb || 60 || 16 || All GPUs associated with the same CPU socket, connected via PCIe
 |-
-|                      192 || v100l ||  32 || 192000M ||  4 || V100-SXM2 || 70 || 32 || Two GPUs per CPU socket; all GPUs connected via NVLink
+|                      192 || v100l ||  32 || 192000M ||  4 || V100-32gb || 70 || 32 || Two GPUs per CPU socket; all GPUs connected via NVLink and SXM2
 |-
-| rowspan=5|Graham  || 160 ||  p100 ||  32 || 127518M ||  2 || P100-PCIE || 60 || 12 || One GPU per CPU socket
+| rowspan=5|Graham  || 160 ||  p100 ||  32 || 127518M ||  2 || P100-12gb || 60 || 12 || One GPU per CPU socket, connected via PCIe
 |-
-|                      7   || v100  ||  28 || 183105M ||  8 || V100-PCIE || 70 || 16 || See [[Graham#Volta_GPU_nodes_on_Graham|Graham: Volta GPU nodes]]
+|                      7   || v100     ||  28 || 183105M ||  8 || V100-16gb || 70 || 16 || See [[Graham#Volta_GPU_nodes_on_Graham|Graham: Volta GPU nodes]]
 |-
-|                      2   || v100(**) ||  28 || 183105M ||  8 || V100-?    || 70 || 32 || See [[Graham#Volta_GPU_nodes_on_Graham|Graham: Volta GPU nodes]]
+|                      2   || v100(**) ||  28 || 183105M ||  8 || V100-32gb || 70 || 32 || See [[Graham#Volta_GPU_nodes_on_Graham|Graham: Volta GPU nodes]]
 |-
-|                      30  ||  t4   ||  44 || 192000M ||  4 || Tesla T4  || 75 || 16 || Two GPUs per CPU socket
+|                      30  ||  t4   ||  44 || 192000M ||  4 || T4-16gb   || 75 || 16 || Two GPUs per CPU socket
 |-
-|                      6   ||  t4   ||  16 || 192000M ||  4 || Tesla T4  || 75 || 16 || &nbsp;
+|                      6   ||  t4   ||  16 || 192000M ||  4 || T4-16gb   || 75 || 16 || &nbsp;
 |-
-| Mist              || 54  || (none) || 32 ||  256GiB ||  4 || V100-SXM2 || 70 || 32 || See [https://docs.scinet.utoronto.ca/index.php/Mist#Specifications Mist specifications]
+| Mist              || 54  || (none) || 32 ||  256GiB ||  4 || V100-32gb || 70 || 32 || See [https://docs.scinet.utoronto.ca/index.php/Mist#Specifications Mist specifications]
 |-
-| Narval            || 159 || a100 || 48 || 510000M ||  4 || A100      || 80 || 40 || Two GPUs per CPU socket; all GPUs connected via NVLink
+| Narval            || 159 || a100   || 48 || 510000M ||  4 || A100-40gb || 80 || 40 || Two GPUs per CPU socket; all GPUs connected via NVLink
 |-
 | Arbutus           ||  colspan=8 | Cloud resources are not schedulable via Slurm. See [[Cloud resources]] for details of available hardware.