Buying Bright Computing gives Nvidia in-house clustering software for high-performance computing and includes support for VMware.
Remember when Nvidia was a gaming-card vendor? That doesn’t seem all that long ago but now it’s a full-blown enterprise high-performance computing and AI company that happens to sell videogame cards – if you can actually find them.
Its latest move is the acquisition of Bright Computing, a maker of Bright Cluster Manager software that controls the configuration of clustered HPC systems, including Nvidia’s own DGX servers and HGX systems made by OEMs and ODMs, plus clusters from other manufacturers. The clusters of servers are linked by high-speed networks into a single unit.
If the deal goes through, Bright Cluster Manager will become a part of Nvidia’s Enterprise Products Group. Nvidia has no intention of keeping Bright Cluster Manager for itself, and by being a part of the Nvidia channel, it gives Bright an opportunity to expand and grow its market.
Bright Computing is a small, privately held firm so details were sparse. Its customer list includes Boeing, NASA, Johns Hopkins University, and Siemens and it serves industries like health care, financial services, and manufacturing.
For Nvidia’s part, it gets an in-house tool to help customers better manage their Nvidia hardware, rather than buy it separately. That’s why it brought the company in-house.
“Nvidia will combine Bright Cluster Manager with our system software capabilities to make HPC data centers easier to buy, build, and operate, creating a much larger future for HPC,” said Charlie Boyle, vice president and general manager of Nvidia DGX systems in a statement. “This will help NVIDIA democratize HPC and accelerated enterprise computing.”
Bright’s software can run in the data center, at the edge, and across multiple public or hybrid clouds. It automates administration for clusters running x86 and Arm as well as Nvidia GPUs. The most recent versions of the software, Bright Cluster Manager 9.1, came out a little more than a year ago, and added support for the entire VMware stack, including the Tanzu Kubernetes container platform.
Bright Cluster Manager 9.1 also added support for Ansible playbooks and integration with OpenShift, so organizations can manage their OpenShift infrastructure from edge to cloud with all of the features Bright offers.
Bright Cluster Manager 9.1 also can automatically increase or decrease the number of nodes available to an HPC workload manager or to Kubernetes in a cluster, regardless of whether those nodes are physical, virtual, on-premises, in the public cloud or at the edge. The allocation of nodes can be determined by demand and by policy.