Zone
SCF
Vendor
Penguin/Dell/NVIDIA
User-Available Nodes
Login Nodes*
1
Batch Nodes
4
Total Nodes
4
CPUs
CPU Architecture
Intel 8480+
Cores/Node
112
Total Cores
448
GPUs
GPU Architecture
H200
Total GPUs
32
GPUs per compute node
8
GPU global memory (GiB)
4512.00
Memory Total (GiB)
4,000
CPU Memory/Node (GiB)
1,000
Peak Performance
Peak PFLOPS (CPUs)
28.400
Peak PFLOPs (GPUs)
272.000
Peak PFLOPS (CPUs+GPUs)
300.400
Clock Speed (GHz)
2.0
OS
TOSS4
Interconnect
Nvidia Mellanox NDR 400
Scheduling Policy (main batch queue)
node-scheduled
Scheduler
Slurm
Recommended location for parallel file space
/p/lustre{...}
Program
ASC
Class
Mixed
Year Commissioned
2026
Compilers

See Compilers page