DIPC Supercomputing Systems Documentation¶
User guides and reference for the DIPC HPC clusters: Hyperion, Atlas EDR, and Atlas FDR.
COMPUTING CLUSTERS
-
Hyperion
17,084 cores across 300 nodes with 220 TB RAM. Features 97 GPUs including NVIDIA A100, RTX 3090, A6000, A30 and RTX PRO 6000 accelerators.
-
Atlas EDR
2,656 cores with EDR InfiniBand interconnect and 16 TB RAM. GPU-enabled with RTX 3090 and P40 accelerators.
-
Atlas FDR
6,116 cores with FDR InfiniBand interconnect and 60 TB RAM. In production since 2015; approaching end of life with minimal maintenance.
USER GUIDES
-
Getting Started
Account setup, SSH access, and your first steps on the HPC systems.
-
Your first job
Step-by-step tutorial to submit, monitor, and evaluate your first real job.
-
Job submission (SLURM)
Full reference: directives, templates, job arrays, partitions, and QoS.
QUICK LINKS
-
Request account
-
Data transfer
-
Script generator
-
Environment modules
-
Software catalog
-
GPU accelerators
-
F.A.Q.