College of Engineering > Administration > Network Services > High Performance Compute Cluster (HPCC) > Network Architecture

Network Architecture

The Samuel Ginn College of Engineering vSMP HPCC consists of 512 cores @ 2.80 GHz X5560, 1.536TB shared memory, and 20.48TB raw internal storage. HPCC Theoretical Performance calculated @ 5.735 teraflops with memory bandwidth up to 90 GB/s.

Note:                                                                             

• Job should be running in any/all of compute nodes, not in head node
• Job can be submitted by a Torque command qsub ./scriptfile.sh
• Torque is a resource manager that sees how many and what resources are available
• Maui is a job scheduler that schedules a job based on available resources informed by Torque
• Sometimes running job in one single node (say with 32 processor) is faster than running job in all four compute node with 8 processor each