Alces Flight is working to make high performance computing broadly available, via solutions built on bare-metal, virtualized, public cloud and hybrid platforms.

There was a time when high performance computing was an exclusive resource whose use was limited to an elite group of scientists and engineers working in university, government and enterprise labs. Not so today, as HPC is now used by even the smallest of startups.

A company that is proving this point is Alces Flight, a U.K.-based firm that serves the HPC needs of many notable Dell Technologies and Intel customers. Alces Flight touts itself as offering HPC for everyone.

For more than 10 years, the U.K.-based company has built integrated solutions, created HPC environments on hardware and cloud platforms, and provided managed open-source HPC projects. The Alces Flight crew creates everything from large-scale, on-premises systems, for use across public and commercial sectors, to scalable cloud systems with fully featured multi-user HPC environments. It serves all sizes of HPC users, from global companies to individual power users.

Want to launch a personal HPC cluster? No problem. Just spin up an HPC cluster in the cloud via the Alces Flight Launch service. This service is designed to get users up and running, quickly and efficiently in a public cloud HPC environment. How quickly? The company says 15 minutes or less.[1]

Rapidly delivering a ready-to-go HPC cluster, complete with job scheduler and applications, the Alces Flight Launch service can be used to perform any HPC task, including application testing and user training. You simply select the HPC cluster you want to evaluate, enter your Flight Launch token from the Alces website, your cluster name, and an email address for notifications and you are ready to go.

Clusters are deployed in a private networked environment for security, and provide SSH terminal access and graphical desktop capabilities for users. Data management tools for POSIX and object storage are also included to help users transfer files and manage storage resources.

In leveraging Alces Flight Launch, users gain the benefit of Alces Flight’s experience with world-class HPC shops. These organizations include several U.K. organizations that are leveraging Alces Flight solutions on Intel-based systems from Dell Technologies.

In one such project, Dell Technologies and Alces Flight worked together to design and build the Barkla HPC cluster at the University of Liverpool. Barkla is a fully managed on-premises HPC cluster based on Dell EMC PowerEdge™ servers with Intel® Xeon® Scalable processors, the Intel® Omni-Path® Architecture (Intel® OPA) and cluster management with Alces Flight Compute.

For a broader and more flexible HPC solution, the Barkla cluster provides a connection to Amazon Web Services (AWS) for additional on-demand public cloud-based computational capacity. With this cloud connection, the University of Liverpool comes closer to its goal of enabling researchers and students to run HPC workloads anywhere, and at any time.

The big picture

With services from companies like Alces Flight and the hardware from companies like Dell Technologies, high performance computing is now broadly available to all sizes organizations, including those that are pushing forward with data analytics, artificial intelligence and other HPC-driven initiatives.

In short, we’ve entered the era of “HPC for everyone.”

To learn more

For a closer look at the Barkla cluster at the University of Liverpool, read the Dell Technologies case study Hybrid HPC.

[1] Depending on size of HPC cluster selected to run within the Flight Launch Service. Average timings indicate a 5-10 minute launch with Standard and Performance Clusters, and a 5-15 minute launch with GPU Clusters.