Credit: Dreamstime
Singapore’s National Supercomputing Centre (NSCC) has awarded Hewlett Packard Enterprise (HPE) a SG$40 million deal to build a new supercomputer for the national high-performance computing (HPC) resource centre.
The new supercomputer to be built under the direct deal will employ the HPE Cray EX supercomputer product line, which is an HPC system designed to support next-generation supercomputing, such as Exascale-class systems. It will also feature a full stack of purpose-built technologies across compute, software, storage and networking.
The centre, which is dedicated to supporting science and engineering computing needs for academic, research and industry, said the new system is expected to provide up to 10 Petaflops of computing capacity and is set to be eight times more powerful than its existing ASPIRE1 supercomputer, which was Singapore’s first national petascale supercomputer.
ASPIRE1, which was commissioned in 2016, has been running at near full capacity in support of local advanced research that requires high-end computing resources. The centre’s existing supercomputer resources consist of 1,288 nodes of CPU and accelerator nodes with NVIDIA K40 GPUs.
The new HPE system, on the other hand, is expected to feature nearly 900 CPU and GPU computing nodes of different configurations, and more than 100,000 computing cores.
Moreover, the NSCC claimed the next-generation national supercomputer for Singapore will be a 'green', warm water-cooled system – one of the first known deployments of such a system in a tropical environment.
The system will also include the HPE Cray EX software stack for software-defined capabilities, which is expected to allow the NSCC’s users to gain the high-performance of a supercomputer, but through the operational experience of a cloud computing delivery model.
Additionally, HPE will integrate a number of next-generation technologies with the HPE Cray EX supercomputer, including expanded storage to support and share complex workloads in modelling, simulation and artificial intelligence (AI) using the Cray ClusterStor E1000 storage system from HPE and HPE Data Management Framework.
Credit: Supplied It will also feature purpose-built HPC networking using HPE Slingshot; a fully integrated software suite to optimise HPC and AI applications using the HPE Cray Programming Environment; powerful compute to improve modelling and simulation using 3rd Gen AMD EPYC processors; and targeted accelerated computing capabilities with 352 NVIDIA A100 Tensor Core GPUs for HPC and AI workloads.
The hardware will be housed in a new data centre designed to increase sustainability and reduce energy consumption, supporting the system's liquid-cooling capabilities made possible through the HPE Cray EX supercomputer line.
The planned HPE-built system is the first in a series that will form the backbone of Singapore’s future supercomputing resources. This will support research in areas like climate change, biomedical science and smart nation activities.
The new hardware is expected to expand and augment ongoing research efforts through the centre by enabling tools such as AI and deep machine learning to optimise modelling, simulation and software simulation for quantum computing.
The deal comes as part of a SG$200 million investment that was announced by the Singapore government in March 2019 to boost the country’s high-performance computing resources.
“We are inspired by how Singapore’s community of scientists have leveraged high performance computing to improve ongoing research efforts,” said Bill Mannel, HPE vice president and general manager. “We are honoured to continue empowering their mission by building them a powerful system using the HPE Cray EX supercomputer that delivers comprehensive, purposely-engineered technologies for demanding research.
“The new system will deliver a significant boost to R&D, allowing Singapore’s community of scientists and engineers to make greater contributions that will unlock innovation, economic value, and overall, strengthen the nation’s position in becoming more digitally-driven,” he added.
The yet-to-be-named supercomputer is expected to be commissioned and operational by early 2022.
