Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Google Cloud Previews HPC VM Image for HPC Workloads

Google Cloud today announced the public preview of HPC VM Image, a CentOS 7-based Virtual Machine Image optimized for HPC workloads with a focus on tightly-coupled MPI workloads, according to Google, making it easier to instantiate VMs that are tuned for optimal CPU and network performance on Google Cloud.

Today’s announcement follows introductions last year of features and best-practice tunings designed for optimal MPI performance on Google Cloud, the company said, demonstrating that MPI ping-pong latency falls into single-digits of microseconds (us) and small MPI messages are delivered in 10us or less. This in turn translates to improved application scaling, expanding the set of HPC workloads that run efficiently on Google Cloud.

But here’s the rub, according to Google’s Pavan Kumar, product manager and Jason Zhang, software engineering manager, who blogged about today’s announcement: building a VM image using these best practices requires systems expertise and knowledge of Google Cloud. Overcoming these complexities, Kumar and Zhang said, by starting with an HPC-optimized image can make it easier to maintain an image.

“The HPC VM image makes it easy and quick to instantiate VMs that are tuned to achieve optimal CPU and network performance on Google Cloud,” they said. “The HPC VM image is available at no additional cost via the Google Cloud Marketplace. “

Google said the HPC VM image is pre-configured and regularly maintained for HPC customers on Google Cloud, offering these features:

  • Create HPC-ready VMs out-of the-box that incorporate best practices for tightly-coupled HPC applications and stay up-to-date with the latest tunings.
  • Networking optimizations for tightly-coupled workloads help reduce latency for small messages, and benefit applications dependent on point-to-point and collective communications.
  • Compute optimizations for HPC workloads allow more predictable single-node high performance by reducing system jitter that can lead to performance variation.
  • Reproducible multi-node performance by using a set of tunings which have been tested across HPC workloads.
  • The HPC VM image is a simple, drop-in replacement for the standard CentOS 7 image.

Google’s announcement included a customer case history involving a research software engineer in the Caltech Particle Theory Group working with the international Bootstrap Collaboration. The collaboration uses SDPB, a semidefinite program solver, to study Quantum Field Theories, with application to a wide variety of problems in theoretical physics, such as early universe inflation, superconductors, quantum Hall fluids, and phase transitions.

To expand the collaboration’s computation capabilities, researcher Walter Landry wanted to see how SDPB would scale on Google Cloud. Working with Omnibond CloudyCluster and leveraging the HPC VM image, Landry achieved comparable performance and scaling to an on-premises cluster at Yale, based on Intel Xeon Gold 6240 processors and Infiniband FDR.

Leave a Comment

*

Resource Links: