NEURONIX   A LOW-COST HIGH-PERFORMANCE CLUSTER

Overview

The Neuronix high performance computing cluster housed in Temple University's Joint Data Center features:

300+ CPU cores		2.6+ terabytes system RAM
20 GPUs		1.5+ petabytes of disk space

We utilize Slurm as a job scheduler, Netdata for live CPU monitoring, and ZFS for disk management. With a maximum theoretical compute performance of 323.54 TFLOPS (SP), Neuronix allows us to conduct big data machine learning research.

Real-Time System Display

node_130 node_000 node_001 node_002 node_003 node_004 node_005 node_006 node_007 node_008 node_011 node_012