NOTE This is a testbed cluster and not for general use.

NOTES:

*1 node: rzwhamo [2]

GPU Peak Performance details: 16.4 TF/s V100s / 7.4 TF/s Mi6; 524.8 TF/s V100s / 236.8 TF/s Mi60

Zone
RZ
Vendor
HPE
User-Available Nodes
Login Nodes*
1
Batch Nodes
32
Debug Nodes
0
Total Nodes
37
CPUs
CPU Architecture
AMD EPYC 7502
Cores/Node
64
Total Cores
2,368
GPUs
GPU Architecture
AMD Mi60
Total GPUs
64
GPUs per compute node
2
GPU peak performance (TFLOP/s double precision)
24.00
GPU global memory (GB)
2.00
Memory Total (GB)
9,472
CPU Memory/Node (GB)
256
Peak Performance
Peak TFLOPS (CPUs)
81.9
Peak TFLOPs (GPUs)
761.6
Peak TFLOPS (CPUs+GPUs)
843.50
Clock Speed (GHz)
2.5
Peak single CPU memory bandwidth (GB/s)
205
OS
TOSS 4
Interconnect
Mellanox HDR / Cray Slingshot
Recommended location for parallel file space
Program
ASC
Password Authentication
OTP
Year Commissioned
2021
Compilers

See Compilers page

Documentation