Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Workload scheduling helps propel Jaguar into Top500

Share this on social media:

The US Department of Energy's Jaguar system, which became the first supercomputer dedicated to open science to break the petaflop processing barrier in the latest Top500 list, is using Moab workload scheduling software from Cluster Resources.

The speed and power of the jaguar aptly characterise the newly achieved petascale capabilities of the Cray XT5 supercomputer system Jaguar at Oak Ridge National Laboratory (ORNL). From the beginning, Moab workload scheduling software from Cluster Resources has been an integral part of ORNL's progression leading to this accomplishment.

'Moab has been essential for getting the most out of Jaguar,' says Don Maxwell, technical lead for Jaguar at Oak Ridge National Laboratory. 'Moab gives us the flexibility to schedule resources to promote leadership science and maintain a very high rate of utilisation. The feature set provided by Moab gives us the capability to change policies easily as requirements and priorities of the center and its users evolve. Development of Moab for the Cray XT platform was initiated in partnership with Cluster Resources three years ago, and we continue to work together to improve the scheduling capabilities for Jaguar.'

As part of this combination of highly sophisticated software and hardware solutions, Moab's predictive scheduling capability optimises utilisation of Jaguar's more than 45,000 quad-core AMD Opteron processors, 360 TB memory, 10 petabyte file system, and other resources. In addition, Moab supports standing reservations, dynamic backfill, preemption, fairshare, quality of service, prioritisation, and 'run-this-job-next' policies.