Building supercomputers is complicated even by the standards of the research institutions and other tech-savvy organizations that most commonly use them. As a result, the implementation process is often incredibly time-consuming. But Hewlett Packard Enterprise believes that it doesn’t necessarily have to be.

The vendor today unveiled a toolkit called Core HPC that attempts to speed up supercomputing projects by bundling the necessary software building blocks into a pre-integrated package. The offering includes a scanner for checking that the hardware components of a deployment are properly installed, a setup wizard designed to streamline configuration and several application services. HPE says that the latter lineup addresses the requirements of both the programmers and administrators who interact with high-performance computing environments.

During the development stage, software engineers use compilers and libraries included in the bundle to optimize their code for their organization’s supercomputer. And once the application is in production, Core HPC’s cluster management tools enables administrators to automate many of the most time-consuming tasks involved in day-to-day operations. The suite packs a scheduler that handles work distribution, a remote control application and monitoring functionality meant to help ease troubleshooting, among other features.

Core HPC should go a long towards strengthening HPE’s presence in the supercomputing market. The company already has a leadership position thanks to its Apollo servers, which have a 35.9 market share within the segment and received an upgrade too as part of today’s rollout. Customers can now order their machines with Xeon Phi coprocessors that each provide an up to 1.2 teraflop speed boost for some workloads. Added up across the hundreds or thousands of nodes in the average HPC cluster, that has the potential to translate into a major performance boost.

The new chip option is complemented by support for Mellanox Technologies Ltd’s speedy 100Gb/s InfiniBand switches, an enhancement geared towards supercomputers with particularly high bandwidth requirements. The hardware integrations and HPC Core are both available immediately.

Image via Wikimedia