A new IDC Technology Spotlight takes a close look at Intel® Scalable System Framework, an architectural philosophy for HPC systems designed not only to address the issues of balance, power efficiency, and reliability in these increasingly powerful machines, but also to provide a common architectural blueprint across a wide range of workloads.
In An Approach for Designing HPC Systems with Better Balance and Performance, IDC explains that, “to move performance on end-user applications forward will require the HPC community to address two key challenges: the imbalanced architectures of today’s HPC systems and the complex and growing set of requirements users have for these systems.”
Intel’s Hugo Saleh, Director of Marketing, Intel High Performance Computing Platform Group, recently explained how Intel Scalable System Framework (Intel® SSF) will address those key challenges.
“System integrators and users have for years designed their HPC clusters around the workloads they will run, which has led to a large diversity of architectures, and it limits others within the organizations from efficiently running their own workloads on the same system,” according to Saleh. “That limits the organizations capability on how they can benefit from their investments or forces them down the path of having multiple clusters, which drives up the costs.” Intel’s vision is to help create systems that converge HPC, Big Data, machine learning, and visualization workloads, along with HPC in the cloud, on a more common framework. “Our goal with Intel SSF is to create systems that can run many types of workloads efficiently while easing procurement and deployment, making HPC resources available to more users, and leading to greater opportunities for discovery and insight,” adds Saleh.
According to Saleh, the design strategy behind Intel SSF focuses on addressing the power, memory, storage, and I/O walls that system builders and operators have run into over the years—but they are also working on the software challenges, which are often part of getting a system up and running. So, architects and developers are taking a three-pronged approach to solving the challenges seen in HPC. “A new, more holistic approach is needed,” says Saleh. “It includes new breakthrough technologies, a new architectural approach to assembling and deploying those technologies, and a versatile and cohesive set of HPC system software stacks that can extract the full potential of the system. We believe that without this system-level approach, ongoing scientific and industrial discovery and innovation upon which so many users depend on HPC advances would be hindered.”
Apparently, system builders like what they are hearing about the new architecture. Leading system integrators and OEMs have already announced launching systems based on Intel SSF, including Colfax, Cray, Dell, Fujitsu Ltd., HPE, Inspur, Lenovo, Penguin Computing, SGI, Sugon, and Supermicro. The Aurora supercomputer, at the Argonne Leadership Computing Facility at Argonne National Lab, will be built on Intel SSF.
Intel’s new architectural framework comprises innovative new and future technologies—such as Intel® Omni-Path Architecture (Intel® OPA), upcoming Intel® Optane™ SSDs built on 3D XPoint™ technology, and future Intel® Silicon Photonics—and incorporates Intel’s existing and next-generation of compute and storage products, including the Intel® Xeon® processors, Intel® Xeon Phi™ processors, and Intel® Enterprise Edition for Lustre* software, Saleh said. Intel OPA is available now, and Intel® Optane™ SSDs will be available in the future.
While Intel is creating breakthrough hardware for HPC, it is also helping enable the software community to take advantage of HPC more easily. Intel has joined with more than 30 other companies to found the OpenHPC Collaborative Project (www.openhpc.community), a new community-led organization focused on developing a comprehensive and cohesive open source HPC software stack. “Anyone who has worked with a team to build out an HPC system knows that it’s a huge challenge to integrate the individual components and validate an aggregate HPC software stack,” states Saleh. “And, it’s a major logistical effort to maintain all the codes, stay abreast of rapid release cycles, and address user requests for compatibility with additional software packages. With many users and system makers creating and maintaining their own stacks, there is duplicated work and inefficiency across the ecosystem. OpenHPC intends to help minimize these issues and improve the efficiency across the ecosystem.”
Download An Approach for Designing HPC Systems with Better Balance and Performance to see why IDC concludes that Intel SSF “aims to provide a foundational direction for building HPC systems that deliver adequate balance, flexibility, and sustained performance for the challenging HPC workloads of today and tomorrow.”