Hardware

SLURM PartitionSLURM GPU
Resource:Type:Numberof
ProcessorNodesTotal Threads +Memory/Core Memory/HostGPU
cpu_gce2.6GHz Dual CPU Eight Core Intel90168GB128GBNone
cpu_gce_test2.6GHz Dual CPU Eight Core Intel10168GB128GBNone
gpu_gcegpu:k40:42.6GHz Dual CPU Eight Core Intel6168GB128GB4x NVIDIA Kepler K40
gpu:p100nvlink:22.4GHz Dual CPU Fourteen Core Intel 1562GB112GB2x NVIDIA P100 (w/NVLINK)
gpu:p100:81.7GHz Dual CPU Eight Core Intel11648GB768GB8x NVIDIA P100
gpu:v100:22.5GHz Dual CPU Twenty Core Intel4404.7GB188GB2x NVIDIA V100
gpu_gce_ppcgpu:v100nvlinkppc64:43.8GHz Dual CPU Sixteen Core IBM Power9 ++11288.5GB1088GB2x NVIDIA V100 (w/NVLINK)
gpu_ose +++gpu:k40:42.6GHz Dual CPU Eight Core Intel21168GB128GB4x NVIDIA Kepler K40
knl_gce1.30GHz Single CPU 64 Core Intel Xeon Phi12560.4GB102GBNone
TOTAL1352,64819TB128

+ In some cases Hyper-threading when enabled may show more threads than physical cores.

++ This worker node is similar to a SUMMIT worker node. 

+++ This partition restricted to Open Science Grid jobs only.