By: Donna Loveland
Software. Most likely, something else comes to mind when you hear the Intel name.
In fact, if you’re in High-Performance Computing, software itself may be something you’d rather not think about. In a recent IDC study (reported June 2016), 29 percent of HPC users identified updating and using system software as “the MOST serious pain point” for their organization. That’s half again as many as those who singled out the heterogeneity of system hardware.
Pleas for consistency have been around the computer industry throughout its history. And no wonder. When everyone can change the hardware ingredients and software components of an underlying platform, developers are basically walking in the dark. Many ISVs find themselves unable to replicate issues yet being charged with solving them, debugging them, and delivering updates. The accelerating adoption of HPC has only aggravated the problem for OEMs and system integrators as well.
Historic approaches – individual companies driving standards or heading standards consortiums – make little sense in HPC. No one, whether you’re a TOP50 lab or turnkey user, could afford the expense and delay that result from duplication of effort or campaigns to advance competing stacks.
Enter OpenHPC. And Intel.
OpenHPC is an open source community project operating under the Linux Foundation. Launched in November 2015, it aims to arrive at a comprehensive software stack to install, manage, and maintain HPC systems.
In its conversations across the HPC ecosystem, Intel heard a common theme: the need for consistency in the system software stack. Whether the discussion centered on removing roadblocks to development or smoothing the path to exascale, lack of a software platform to accommodate individual and group-level innovation emerged time and again as a root cause.
When Intel made a case for community action last year through the Linux Foundation it opened the floodgates to this unmet need. OpenHPC’s inaugural gathering drew participants from across the globe and every area of the ecosystem. They included software vendors, research institutions, equipment manufacturers, supercomputing sites, and end users.
The OpenHPC software stack includes provisioning tools, I/O clients, development tools, and a variety of scientific libraries. The Community pulls relevant, reliable, validated versions of system management components from upstream communities and acts as a midstream liaison to integrate a stack that becomes the base for HPC system management software.
To kick-start the work, Intel seeded the Community with an integrated stack that OpenHPC demoed at SC15 and at ISC 2016. It’s being embraced, downloaded at the rate of hundreds per month.
In the near term, Intel is addressing the practical issue of taking OpenHPC into production-ready environments by launching a product it calls Intel® HPC Orchestrator.
Announced during ISC 2016, HPC Orchestrator is the Intel distribution of the OpenHPC stack, optimized for Intel hardware ingredients. Recognizing that every business needs to keep its operations running smoothly, Intel backs HPC Orchestrator with professional services and support – what you need whether you’re a manufacturer who can’t lose a day to breakdowns or a shop too small to find quick help with homegrown applications.
As the software portion of the Intel® Scalable System Framework, HPC Orchestrator is the glue that connects the goodness in the hardware to the software. It’s pre-integrated, pre-tested, and pre-validated.
The OpenHPC stack, backed by HPC Orchestrator, can turn pain to gain. Innovations shared through the Community can be implemented with confidence of success. Together, the community will work to boost system performance to exascale, a rising tide that will lift all boats across the ecosystem.