Research
Machines

Quest High Performance Computing Cluster Nodes

The CS department has partnered with Northwestern IT Research Computing to provide priority access to 4 nodes (computers) on the Quest HPCC. These nodes are colloquially referred to as the “CS Quest Enclave.”

The pilot enclave includes:

  • 10 TB storage space
  • 2 x GPU nodes/computers, each with:
    • 4 x NVIDIA A100 GPUs (80 GB SXM4)
    • 2 x Intel Xeon Gold 6338 32-core CPUs
    • 512 GB memory
  • 2 x CPU nodes/computers, each with:
    • 2 x Intel Xeon Gold 6338 32-core CPUs
    • 256 GB memory

Local CS GPU Servers

The following is a list of CS-purchased GPU servers arranged in two groups: systems with a VM-based allocation scheme and systems configured to use the SLURM scheduler across their nodes (like Quest). These systems are primarily used for faculty-directed machine learning research projects.

VM-Based Systems

athena, glamdring, sting

Each with:

  • 8 x NVIDIA Quadro RTX 8000 48 GB GPUs
  • 2 x Intel Xeon Gold 5218 16-core CPUs
  • 384 GB memory

SLURM-Based Systems

nyx, hemera, erebus

Each with:

  • 4 x NVIDIA Ampere A40 48 GB GPUs
  • 2 x Intel Xeon Gold 5318S 24-core CPUs
  • 384 GB memory

CS CPU Server

Primarily used for faculty-directed research projects.

osprey.cs.northwestern.edu

  • 2 x Intel Xeon Gold 6140 18-core CPUs
  • 384 GB memory

Requesting Access

The above resources are available to the Northwestern CS community, enabling faculty and students to conduct research in a variety of computer science areas. If you are interested in using these machines, please have your faculty advisor contact Professor Bryan Pardo to apply for access.