Stay informed with free updates
Simply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.
The head of the £800mn supercomputer project cancelled by the Labour government has warned the UK risks stifling “science and innovation” by not investing in the advanced technology.
Mark Parsons, supercomputing professor at Edinburgh university whose exascale venture had its funding scrapped over the summer, said it would be a “disaster” if the UK did not restart efforts to build next-generation capabilities.
“We can’t be a country the scale of Britain without a supercomputer,” he told the Financial Times. “It would block the advancement of UK science and innovation.”
His warning came as figures show the UK’s most powerful computer has been overtaken by rivals, meaning the country no longer has a machine ranked in the global top 50, in a potential setback for AI development.
Restarting Edinburgh’s programme would cost £840mn over seven years, with no more than £300mn in a single year, while the project would become operational within two years, Parsons said.
“I recognise that is a lot of money, but it is a fraction of what our international peers are investing in supercomputing and AI,” he added.
The government came under fire in August for scrapping exascale’s funding, although tech secretary Peter Kyle insisted he did “not cut anything” because the £800mn promised by the previous government was not budgeted for.
Parsons blamed a “lack of long-term thinking by successive governments”, and said the government “needs to make some decisions quite quickly”.
He added: “I don’t think it’s properly understood how much we rely upon modelling and simulating the world around us,” with modern planes and cars relying on digital modelling during their development.
Exascale supercomputing — defined as the ability to produce a billion billion operations a second — is widely seen as a crucial step to unlocking the widespread adoption of AI. There are only three known fully functional exascale computers in the world, all in the US, though six are expected to be online by the end of next year.
China does not make all of its compute capacity public, although experts believe it has at least one exascale in operation.
Researchers and tech executives had hoped money for supercomputing would be committed in the Autumn Budget last month, though this did not happen. Now many are pinning their hopes on the spending round in the spring, where the future of exascale in Edinburgh could be salvaged.
Calls from senior scientists and business figures to move with urgency in financing compute in Britain come as data released on Monday by respected computing ranking index The Top 500 showed the UK had dropped out of the ranking of the top 50 most powerful computers in the world.
Parsons also rebutted the claim made by some in the sector that Edinburgh’s exascale programme was not geared enough towards artificial intelligence applications, and would have been more focused on scientific modelling.
“AI systems . . . and simulation systems aren’t in conflict with each other — they can do both. Japan, the USA, Germany, China — all of these countries are doing both,” he said.
Archer2, the supercomputer programme already being run out of Edinburgh university that is due to reach its end of life in 2026, is “fundamental to how we do science today”, Parsons said, adding that it has 3,000 active users across science and business.
His concerns were echoed by Hewlett Packard Enterprise director Matt Harris, whose business provides the technology for most of the leading supercomputers in the west.
Replacing Archer2 with another competitive system was a “national imperative”, he said, warning: “If you compare us to every other developed nation, the danger is we’re falling behind.”
It would be an error for the government to rely on the private sector to replace state investment because of the vast scale of compute need, he said.
“Taking a long time on any decision is dangerous for [Britain’s] economic viability,” he said, adding that the government’s “failure to invest is a concern”.
The government did not immediately respond to a request for comment.
Read the full article here