01 Dec 2022

Recommendations for UK- Japan collaboration for the Future of Compute

On 9 November, techUK hosted a virtual roundtable exploring the UK and Japanese approaches to building compute infrastructures, and explore R&D and industry collaboration opportunities

Future of Compute week generic card.png 1

On 9 November, techUK hosted a virtual roundtable exploring the UK and Japanese approaches to building compute infrastructures, and explore R&D and industry collaboration opportunities.

In the session we addressed how UK and Japan can leverage bilateral cooperation to broaden accessibility of HPC, large-scale computer systems and quantum systems to SMEs and what mechanisms and initiatives both governments and industry need to put in place to drive bilateral industry to industry exchanges.

Speakers included:

  • Hisashi Kanazashi, Director, IT Industry Division, METI
  • Jenny Hall, Co-Director for International, DCMS
  • Andrew Dean, Sales Director, OCF Limited
  • Dr Dave Snelling, Director of Advanced Compute, Fujitsu
  • Hirochika Asai, VP of Infrastructure Strategy, Preferred Networks, Inc.

Why does UK-Japan collaboration in future of compute matter?

High performance and other forms of large-scale compute are a key enabler of science, research and innovation and sits at the heart of value chains that extend deep into the economy. High performance computing is essential to R&D in a range of sectors, including healthcare and drug discovery, engineering and manufacturing, financial services and risk analysis, particle physics, materials science and more. It is viewed as a key part of countries’ ambition to remain relevant in science and innovation. Countries around the world are increasingly looking to develop their capabilities in this area to not fall behind competition.

Japan has recently pledged to invest heavily in its compute infrastructure as it is seen vital for maintaining and improving Japan's scientific and industrial capabilities. The sixth Science, Technology and Innovation plan (FY 2021 – 2025) - building on the Society 5.0 plan - further emphasises the importance of investing in R&D for next generation infrastructure including high performance computing and quantum technology.

Equally, the UK Government, led by DCMS, have announced the upcoming future of compute strategy to understand what the UK’s advanced compute needs will be in the next decade, and how government should meet them. techUK believe compute infrastructure is strategic national infrastructure, as important to our economic future in an information age as steel was in the industrial age. 

While Japan is one of the world’s leaders in both power and total capacity of high-performance compute, the UK holds global leadership in data analytics, cybersecurity, AI, machine learning and quantum technologies. As there is convergence of compute infrastructure with other key technologies like quantum, cloud, and AI, there are collaboration opportunities for both countries in driving future of compute forward.

During the discussion, the panellists identified a number of challenges to advancing compute infrastructure, including:

  • Resilient supply chains providing secure access to advanced chips, semiconductors and quantum computers. Supply chain sustainability and resilience are tier one global issues in the post-pandemic world. In 2020, the continuous stream of global trade ground to a halt. The resulting commodity prices increases and widespread part and product shortages made clear that if one link in the chain is broken, the knock-on effects are significant.
  • Availability of skills and talent. A shortage of talent can hinder the development of compute infrastructure and slow down SMEs access to high-performance computing. Insufficient supply of skilled professionals will lead to fierce international competition for scientists, engineers and software developers, further inhibiting the progress in the field.
  • Lack of knowledge sharing, R&D partnerships and industry collaborations. International partnerships on industry and academia level are critical in this space. They facilitate knowledge exchange and build long-lasting industry and academia links driving the scientific and commercial aspects of the field forward.
  • Energy Efficiency. Supercomputers are energy demanding which raises concerns about rising energy prices and net-zero targets. International collaboration with other nations with cool climates could be one viable solution here
  • Data sharing. Sharing anonymised data for public sector research is essential to make the most effective use of limited resources. However, sharing data safely and effectively must take account of privacy concerns. Divergent approaches to data regulations and privacy concerns pose restrictions to data sharing across countries. This obstacle affects the transfer of data to foreign institutions and also remote access by other researchers to data at its original location. Both activities are essential for international collaborative research.
  • Lack of standardisation and interoperability. Interoperable regulatory regimes and harmonised standards make it easier to run open global networks and next generation of compute infrastructure systems as businesses can operate in overseas jurisdictions without having to adapt their products to comply with local regulations and standards.

Below are some of the key messages for future UK-Japan collaboration in the future of compute:

  • The UK and Japan should pursue bilateral supply chain cooperation, especially on strengthening and diversifying semiconductor production capacity, promoting workforce development, increasing transparency, coordinating emergency response to semiconductor shortages, and enhancing R&D cooperation. The terms of cooperation should be formalized in a joint statement or MoU. Both countries should launch a formal dialogue to identify gaps in the global supply chain and lay out a roadmap to address them, including through joint research and investment projects. In the mid- to long-term, they should cooperate to increase capacity across the entire semiconductor value chain and bring about a more geopolitically balanced production.
  • Skills development within industry and cross-border cooperation on the development of bilateral and national educational programs are strongly encouraged. National professional organizations should work together to develop programmes aimed at providing new skills to existing workers who want to get into the quantum and HPC field. Both countries should focus on developing programmes facilitating academia exchanges  and skills – joint training, staff exchange.
  • The UK and Japan are encouraged to build upon commitments on regulatory cooperation made in the UK-Japan CEPA. The regular engagement between regulators, government and industry stakeholders from both countries is crucial to developing common regulatory approaches to governing emerging technologies supporting the next generation of compute infrastructure. Businesses need to be actively involved in informing the recently established UK-Japan Digital Group as well as the UK-Japan Science and Innovation Committee.
  • Both countries have complementary capabilities in the supercomputing environment. It was suggested to build a strategic platform between Japan and the UK which could serve as a “sandbox”. This would allow both countries to test ideas and different models and develop a next generation platform built on net-zero principles.
  • The UK and Japan have already sought to lead international efforts here by agreeing a G7 roadmap for Cooperation on Data Free Flow with Trust (DFFT). We invite the UK and Japan to continue to lead the international discussion on the importance of policy coordination on data governance and flows as well as new technologies supporting data processing such as AI. Both the UK and Japan can also take the lead internationally by advocating for the free flow of data across the borders of likeminded countries and taking a firm stance against countries who already have or plan to implement data localisation policies.
Laura Foster

Laura Foster

Head of Technology and Innovation, techUK