The Lewis HPC cluster is a shared cluster located in Columbia Missouri accessible to researchers and their collaborators at the University of Missouri and across the University of Missouri System. Lewis currently runs CentOS 7 and uses SLURM to schedule jobs. It currently consists of 11 Haswell nodes with 24 cores and 128GB and 72 Broadwell nodes with 28 cores and 256GB of RAM connected with 10 Gigabit Ethernet and 40 Gigabit QDR Infiniband and 32 Broadwell nodes with 56 cores and 512GB of RAM connected with 10 Gigabit Ethernet. It also consists of 35 Skylake nodes with 40 cores and 384GB of RAM connected with 25 Gigabit Ethernet and 100 Gigabit EDR Infiniband and 64 nodes with dual Cascade Lake processors with 48 cores and 384GB of RAM per node connected with 10 Gigabit Ethernet and 100Gbps HDR Infiniband for a total of 8544 cores. There are also 144 Sandy Bridge cores for interactive jobs and testing. There are currently 14 GPU nodes with 6 NVIDIA K20m GPU’s, 1 NVIDIA K40, 1 NVIDIA P100, 7 GTX 1080Tis, and 12 NVIDIA v100 for a total of 27 GPU’s. The system is connected to 925TB of shared high speed parallel HPC storage (Lustre) for computation and 2040TB of HTC storage for large economical low computational intensity project storage (ZFS). The cluster is supported in part by an NSF MRI Award 1429294.
For more information on how to write Lewis into your grant proposals, please send an inquiry to firstname.lastname@example.org.