Difference between revisions of "CSLT cluster queues"

From cslt Wiki
Jump to: navigation, search
Line 11: Line 11:
 
|-
 
|-
 
|wolf.gpu.q || wolf03=1, wolf04=1, wolf05=1, wolf06=1, wolf07(grid-19)=2, wolf08=2 || GTX 970 * 8 ||
 
|wolf.gpu.q || wolf03=1, wolf04=1, wolf05=1, wolf06=1, wolf07(grid-19)=2, wolf08=2 || GTX 970 * 8 ||
 +
|-
 +
|tiger.cpu.q || tiger01=8 || e3 1230v5, 8 cores ||
 +
|-
 +
|tiger.gpu.q || tiger01=2 || GTX 1070 * 2 ||
 
|-
 
|-
 
|}
 
|}

Revision as of 01:32, 16 August 2016

  • Now we have several queues. Info of these queues are below:
Queue Node & Slot Processor
sheep.cpu.q sheep02=8, sheep03=8, sheep04=8 i7, 8 cores*3 = 24
sheep.gpu.q sheep02=1, sheep03=1, sheep04=1 GTX560Ti * 3
wolf.cpu.q wolf01=8, wolf02=8, wolf03=8, wolf04=8, wolf05=8, wolf06=8, wolf07(grid-19)=8, wolf08=8 i7(or e3), 8 cores*8 = 64
wolf.gpu.q wolf03=1, wolf04=1, wolf05=1, wolf06=1, wolf07(grid-19)=2, wolf08=2 GTX 970 * 8
tiger.cpu.q tiger01=8 e3 1230v5, 8 cores
tiger.gpu.q tiger01=2 GTX 1070 * 2