The N8 Centre of Excellence in Computationally Intensive Research, N8 CIR, has been awarded £3.1m from the Engineering and Physical Sciences Resources Council to establish a new Tier 2 computing facility in the north of England. This investment will be matched by £5.3m from the eight universities in the N8 Research Partnership which will fund operational costs and dedicated research software engineering support.
The new facility, known as the Northern Intensive Computing Environment or NICE, will be housed at Durham University and co-located with the existing STFC DiRAC Memory Intensive National Supercomputing Facility. NICE will be based on the same technology that is used in current world-leading supercomputers and will extend the capability of accelerated computing. The technology has been chosen to combine experimental, modelling and machine learning approaches and to bring these specialist communities together to address new research challenges.
NICE will make it easier to use machine learning alongside traditional modeling and simulation to better understand the vast datasets that are generated by experimentalists at national facilities such as Diamond and the Henry Royce Institute or international facilities including the European Synchrotron Radiation Facility.
Dr Alan Real, Technical Director of N8 CIR and Director of NICE said, ‘This supercomputer will increase the diversity of platforms available to the UK research community, expanding data sizes for accelerated simulation and analysis codes, and providing a route to exascale computing. Investments in STFC’s IRIS will be leveraged across the partnership and beyond to connect the facility to experimental apparatus for optimal data flows.’
NICE will comprise 32 IBM Power 9 dual-CPU nodes, each with 4 NVIDIA V100 GPUs and high performance interconnect. This is the same architecture as the US government’s Summit and Sierra supercomputers which occupied the top two places in a recently published list of the world’s fastest supercomputers. An additional 6 nodes will use T4 and FPGA technology targeted towards AI inference, to improve error estimation and robust prediction. This architecture supports memory coherence between the GPU and CPU and a hierarchy of interconnects to allow effective distributed GPU use, extending problem sizes that can be tackled beyond that of other GPU-accelerated architectures.
The N8 partnership is supporting all operational costs including dedicated research software engineering support at each institution to help with user training, code porting and optimization, to maximize the benefit of this new computing resource.
Through the N8 CIR working in partnership with IBM, OCF and NVIDIA, the new centre will contribute to EPSRCs’ and N8’s goals through:
- Engaging with new and existing research communities
- Delivering a new computing architecture with new technologies
- Preparing the UK for new Tier-1 and Tier-0 architectures
- Accelerating the ‘time to science’ for a range of ‘hard’ problems
- Integration into the existing Tier-2 and National e-infrastructure
- Helping develop new computational skills for the RSEs and the researchers
- Developing new links with high-profile international supercomputing centres
- Enabling multi-disciplinary science through its unique architecture
