Dynamic Resources for HPC Consortium

What we do

The Dynamic Resources for HPC Consortium advances dynamic resource management, malleable applications, and adaptive scheduling for High-Performance Computing. We bring together experts and practitioners from industry, academia, national labs, and HPC centers worldwide through monthly meetings, workshops at major conferences, and technical working groups.

DynResHPC shares knowledge via events, joint initiatives, and community resources. Join our community and attend virtual meetings on the 3rd Monday of each month at 5:00 PM CET. If you would like to participate, see how to join or subscribe to our mailing list. Upcoming and past events are listed in the sidebar and on the News page.

Frequently asked questions

What is Dynamic Resource Management?

Dynamic Resource Management (DRM) is primarily about how HPC clusters and supercomputers assign compute resources to jobs. In a typical batch system (for example Slurm or similar resource managers), an application receives a fixed set of nodes, cores, or GPUs at submission time and keeps them until it finishes. DRM instead allows those allocations to grow or shrink while the job is running, so malleable applications and adaptive schedulers can improve throughput, utilization, and energy use on shared clusters.

What is DynResHPC?

The Dynamic Resources for HPC Consortium is a collaborative group focused on advancing dynamic resource management, malleable applications, and adaptive scheduling in High-Performance Computing (HPC) systems.

How did it get started?

The consortium originated from a series of informal discussions and small-scale collaborative events such as hackathons. During SC24, we held the first official meeting of the consortium, marking the beginning of a coordinated community effort. Since then, we have been meeting monthly to share progress, exchange ideas, and plan joint activities.

Who is part of DynResHPC?

Anyone interested in dynamic resource management in HPC — scheduling, runtimes, applications, APIs, monitoring, simulators, modeling, and related topics. Participants come from industry, academia, national labs, and HPC centers around the world.

What events and meetings does DynResHPC hold?

We hold monthly virtual meetings on the 3rd Monday of each month, plus in-person events at major conferences such as SC and ISC. See the News page for upcoming and past events.