Altair
472 Case Studies
A Altair Case Study
GE’s Oil & Gas engineering group (formerly Nuovo Pignone) needed to get a newly upgraded HPC cluster into production to run commercial and proprietary CFD/CEA tools such as StarCD, ANSYS CFX, Fluent and GE’s Tacoma. After moving from an 8‑CPU UNIX server to a 20‑CPU HP ProLiant cluster they encountered persistent job‑scheduling failures with open‑source solutions, so they engaged Altair to resolve the urgent workload‑management problem using PBS Professional.
Altair deployed and customized PBS Professional, tuned the cluster and Gigabit Ethernet configuration, and integrated the major applications; GE then purchased PBS Professional licenses for the entire cluster. The result was a production‑ready environment that scaled to 45 two‑processor nodes (90 processors, including 14 64‑bit nodes), supported 15–20 concurrent engineers running large 6‑million‑cell models across 12–16 processors, and enabled efficient job management with plans for cycle harvesting and cross‑group resource sharing.
Alessandro Ciani
Group Leader, Heat Transfer and Secondary Flows