The Turing high-performance computing (HPC) cluster is intended for research-intensive computational science — specifically applications able to make use of its special capabilities such as distributed-memory computation, multi-CPU SMP parallelization, or GPU acceleration. Users who cannot or do not intend to use these capabilities should use other systems.
6,750
CPU cores
180+
GPUs
64 TB
Memory
1.5 PB
Ceph storage
Hardware
| CPUs | 6,750 cores — Intel Scalable Platinum 8168, Intel Xeon 6767P, AMD EPYC 9654 |
|---|---|
| Memory | 64 TB aggregate |
| GPUs | 180+ accelerators, including 40× RTX Pro 6000 Blackwell and 16× NVIDIA H200 |
| Interconnect | 100 Gigabit Ethernet internal · 10 GbE to the WPI campus |
| Storage | 1.5 PB high-performance Ceph scale-out system — home directories & scratch, with hourly self-service snapshots |
Software Environment
| Operating system | Ubuntu 20.04 |
|---|---|
| Cluster management | Bright Cluster Manager |
| Job scheduler | SLURM |
| Parallel libraries | MPI, PETSc, LAPACK, BLAS, and other HPC software |
| Applications | MATLAB, Ansys, Abaqus, Fluent, COMSOL, Intel & PGI compilers, and more |
Not sure Turing is the right fit? If your workload doesn't use
distributed-memory, SMP, or GPU parallelism, ARC staff can point you to a better-suited
resource. Schedule a consultation.
Requesting Access
Each Turing account request is evaluated on an individual basis. Submit the Turing account request and include details on the software you will be using. For Turing cluster support, contact gr-arcunix@wpi.edu.
Part of WPI Academic & Research Computing