Bristol to host national ARM supercomputer

The Catalyst UK project is led by HP Enterprises with chip designer ARM, software provider SUSE and the universities of Bristol, Edinburgh and Leicester. It will see the development and use of one of the largest ARM-based supercomputing installations worldwide, allowing companies and academics to build new applications that use vast amounts of data.

• You may like: Bristol and Bath to build world’s largest ARM-based supercomputer

The supercomputer installations will be spread across the three universities this summer and the project will run for three years.

“Bristol’s early experience with Arm via the EPSRC-funded GW4 Isambard project, and the European FP7-funded Mont-Blanc 2 project, gave us the confidence to explore deploying Arm-based supercomputers for real workloads in a production environment,” said Simon McIntosh-Smith, Bristol Professor of High Performance Computing and Head of the HPC Research Group at the University of Bristol. “Bristol has a wealth of experience porting and optimising HPC applications for the Arm architecture in general, and Cavium ThunderX2 processors in particular. Through this new Catalyst UK programme, the HPE Apollo 70 HPC systems will, for the first time, enable us to apply that experience to explore scaling across InfiniBand [a high-speed networking technology]. We expect these results to be of great interest to our industrial and academic partners.”

At Bristol, the new Arm system will complement the existing Isambard GW4 Tier 2 HPC service. Isambard, the world’s first production Arm-based supercomputer, is funded by the Engineering and Physical Sciences Research Council (EPSRC) and run jointly by the GW4 universities of Bristol, Bath, Exeter and Cardiff, along with the UK’s Met Office. This new project will enable Bristol to expand its focus on ARM-based supercomputing, making it one of the first sites anywhere in the world to run multiple ARM systems from multiple suppliers.

ARM-based supercomputers are one of the current approaches to overcome the limitations of traditional computer architectures and offer a better price-performance ratio for modern workloads and applications. This includes AI, which needs to process large amounts of data and requires extremely high memory bandwidth, and exascale computing, which requires HPC systems to be hundreds of times faster and more efficient than today’s fastest supercomputers, capable of a billion billion calculations per second.

The three supercomputer clusters at EPCC, University of Bristol and the University of Leicester will in total run more than 12,000 Arm-based cores, hosted by HPE Apollo 70 HPC systems. The clusters at each university will be largely identical, consisting of 64 HPE Apollo 70 systems, each equipped with two 32 core Cavium ThunderX2 processors, 128GB of memory composed of 16 DDR4 DIMMs with Mellanox InfiniBand interconnects. The operating system is SUSE Linux Enterprise Server for HPC. Each cluster is expected to occupy two computer racks and consume a total of approximately 30KW of power.

“We are currently seeing an insatiable demand for compute performance, as companies seek to gain intelligent and actionable insights from their data. As we embark on the global race towards more powerful and eventually exascale systems, new approaches and technologies are needed to tackle some of the key challenges in achieving these levels of performance, such as rising energy consumption,” said Mike Vildibill, VP, Advanced Technologies Group at Hewlett Packard Enterprise.

Researchers at Bristol also have access to mainstream supercomputers for research at the High-Performance Computing (HPC) service with BrisSynBio

• You may like: World’s first quantum encryption chip developed