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


Red Hat’s Relentless Focus on Open Source Software Challenges Status Quo in Supercomputing

In this guest article, Red Hat Chief Technology Officer Chris Wright explores how his company’s focus on open source software is challenging the status quo in HPC and supercomputing. 

open source

Chris Wright, CTO, Red Hat

Supercomputers and scientific research tend to go hand-in-hand. Designed for solving fundamental scientific problems, supercomputers have increasingly become more open in terms of collaboration and information exchange. In the past, supercomputer designs and implementations were relatively closed, often built by a single vendor from their inventory of components. Today, this model is evolving: As science becomes more accessible through global community, so to are supercomputers opening up via vendor collaboration.  

Two recent examples are the Summit and Sierra supercomputers rated #1 and #3 in the world, respectively. The result of multi-year cross-vendor collaboration efforts, these systems are using CPUs from IBM, GPU accelerators from NVIDIA, Infiniband networking from Mellanox, and a Linux operating system from Red Hat in Red Hat Enterprise Linux. This highlights a new path for not just supercomputing, but enterprise computing: more seamless multi-architectural support. Access to a broader range of architectural choices enables organizations to choose the computing backbone that best meets their unique needs, whether it’s a traditional datacenter environment or a high-powered supercomputer. 

 The compute resources required by supercomputers go well beyond how we would normally talk about flexibility and scalability for IT operations. Supercomputing often pairs standard hardware at scale with additional, highly-specialized components. All of these components are commoditized, in that they can be (and often are) deployed into enterprise datacenters. Linux forms a common bridge to effectively link all pieces together, making it easier for individual application stacks to take advantage of the specific resources that they need. So it should not come as a surprise that, according to the most recent Top500 list, all supercomputers in the world run a variant of Linux.

open source

Providing a consistent environment across all of these footprints is Red Hat Enterprise Linux, a proven platform for large-scale, multi-architecture computing. Many leading researchers around the world are already familiar with the operating system, simplifying their ability to build and execute research jobs at scale.

Another great example of close vendor collaboration for high-performance implementations is the recently announced Astra supercomputer at Sandia National Laboratories. Born out of an early access program from HPE, Cavium, and Red Hat, Astra is based on the new Arm server architecture. Astra is first of a potential series of advanced prototype platforms and will help determine the production feasibility of this emerging high-performance computing architecture.  

In collaboration with our partners, Red Hat is enabling new hardware architectures, supporting various acceleration technologies and cache-coherent interconnects, and driving open innovation and standardization in high-performance computing. Through collaborative community efforts like OpenHPC and industry partnerships such as the OpenPOWER foundation and CCIX, we’re helping to reshape the next generation of supercomputers to be more open and stable and to support the globally-shared nature of scientific advancement. 

All of this, combined with Red Hat’s expertise in supporting open source systems, makes Red Hat portfolio of products a powerful platform for any supercomputing endeavors. 

Chris Wright is the Chief Technology Officer at Red Hat. 

Leave a Comment

*

Resource Links: