Background: UOM CIR Ecosystem

Batch Compute
  • CSF: general-purpose compute cluster (9000 cores).
  • DPSF: mostly 512 GB RAM nodes, large/fast /scratch, dedicated Isilon/RDS.
  • Condor: short, small jobs only (3000 cores).
Also. . .
  • iCSF: for interactive (e.g, GUI-based) work.
  • Research VM Platform
Research Data Storage
  • 3 PB two-tier, replicated and snapped — soon 6TB?
  • 0.5 PB dedicated to DPSF.
  • All connected by dedicated, secure, fast networking.



[Preview] [Continuous/One Page]


Page Contents: