OIT Enhances the University’s Shared High Performance Computing Cluster

May 31, 2017


(Entrance of the UTSA Research Data Center)

By Danicia Steele, Communication Specialist

In summer 2017, the UTSA Office of Information Technology made updates to Shamu. This enhancement took place on Sunday, May 14 by OIT’s Enterprise Network Engineer Matthew Hoke. This update will benefit the University by providing better data transfer rates as well as increased disk performance and capacity.

Shamu is a centrally shared high-performance computing (HPC) cluster currently comprised of:

  • 52 physical servers.
  • 1784 total CPU Cores.
  • 13 TB of shared memory.
  • 96 TB of shared disk storage.
  • 4 Nvidia Tesla K80 GPU nodes.
  • 1 node with 72 E7-8800v4 XEON cores and 1.5 TB RAM.

Shamu has many different software packages available including MATLAB, Abaqus, ParaView, VisIt, Visualization Toolkit (VTK), Octave, JAVA, Mathematica, and LS-DYNA.

Before the update, Shamu only had a network connectivity of 1 Gb/s (Gigabit per second) available, which gave users limited network access. Now users have unlimited access to 10 Gb/s of connection which will enhance access to research resources on campus. Researchers now have the ability to connect to Shamu regardless of what speed their personal device supports.

“Professors that upload extremely large data sets for calculations to be done quickly now will benefit from faster results from the 10 Gb/s,” said Hoke. “We are trying to supply what researchers need today and build a framework for increasing bandwidth and ease of access to research resources.”

If a researcher has a connection with speeds lower than 10 Gb/s, the UTSA Research Computing Support Group (RCSG) provides a set of computing workstations in the UTSA Advanced Visualization Lab (MS 1.03.06N). The workstations are open to UTSA students, faculty and staff and allow users to perform research using 10 Gb/s of speed without any additional cost.

For more information on UTSA Shamu, please click here

Click here for more information on the UTSA Advanced Visualization Lab (VizLab)