Adaptive Computing Leverages TORQUE to Manage Intel Xeon Phi Resources

Print Friendly, PDF & Email

In this video from SC12, Nick Ihli from Adaptive Computing demonstrates how the company’s Torque resource manager works with Intel Xeon Phi. By relaying Intel Xeon Phi instrumentation such as memory availability to the company’s Moab workload manager, the system is able to schedule coprocessor resources efficiently.

With the amazing capabilities of the latest supercomputing coprocessors such as the Intel Xeon Phi coprocessor, it’s vital to make it as simple as possible to integrate them into existing supercomputers,” noted Robert Clyde, CEO of Adaptive Computing. “The latest iteration of Moab was designed to maximize the investment being made by today’s HPC providers.”