Stay up so far with free updates
Simply log in Artificial intelligence myFT Digest – delivered straight to your inbox.
The head of the £800m supercomputer project canceled by the Labor government has warned that the UK risks stifling “science and innovation” if it doesn’t put money into the advanced technology.
Mark Parsons, a supercomputing professor on the University of Edinburgh whose exascale project was canceled over the summer, said it will be a “disaster” if the UK didn’t restart its efforts to construct next-generation skills.
“We can’t be a rustic the scale of Britain with no supercomputer,” he told the Financial Times. “It would block the progress of British science and innovation.”
His warning got here as figures show Britain's strongest computer has been overtaken by rivals, meaning the country not has a pc on the earth's top 50, in a possible setback for the event of artificial intelligence.
Restarting the Edinburgh program would cost £840 million over seven years, with not more than £300 million incurred in a single 12 months, while the project can be operational inside two years, Parsons said.
“I realize that is plenty of money, but it surely is simply a fraction of what our international colleagues are investing in supercomputing and AI,” he added.
The government got here under fire in August for cutting funding for exascale, despite tech secretary Peter Kyle insisting he had “not cut anything” since the £800m promised by the previous government had not been budgeted for.
Parsons blamed this on a “lack of long-term considering on the a part of successive governments” and said the federal government needed to “make some decisions quite quickly”.
He added: “I don't think it's fully understood how much we depend on modeling and simulating the world around us,” as modern planes and cars depend on digital modeling of their development.
Exascale supercomputing – defined as the flexibility to perform one billion billion operations per second – is widely seen as a critical step in enabling the widespread adoption of AI. There are only three known fully functioning exascale computers on the earth, all within the United States, although six are expected to be online by the top of next 12 months.
China doesn’t disclose all of its computing capability, although experts imagine it has at the very least some exascale computing capability in operation.
Researchers and technology executives had hoped last month that cash can be allocated for supercomputing within the autumn budget, but that didn't occur. Now many are pinning their hopes on the spring spending round, which could save the longer term of exascale in Edinburgh.
Leading academics and business leaders are calling for rapid progress to be made on funding computers within the UK after data released on Monday by the respected computer rankings index The Top 500 showed the UK has fallen out of the rankings of the world's 50 strongest computers.
Parsons also refuted claims by some industry figures that Edinburgh's exascale program was not sufficiently focused on AI applications and had focused more on scientific modeling.
“AI systems. . . and simulation systems usually are not contradictory – they will do each. Japan, the US, Germany, China – all these countries are doing each,” he said.
Archer2, the supercomputer program already running on the University of Edinburgh and scheduled to succeed in its end in 2026, is “fundamental to the way in which we do science today,” Parsons said, adding that there are 3,000 energetic users in science and business.
His concerns were shared by Hewlett Packard Enterprise director Matt Harris, whose company supplies the technology for many of the West's leading supercomputers.
Replacing Archer2 with one other competitive system is a “national necessity,” he said, warning: “If you compare us to another industrialized nation, we risk falling behind.”
It can be a mistake for the federal government to depend on the private sector to switch government investment due to huge computing needs, he said.
“It is dangerous for (Britain's) economic viability to take too long on a call,” he said, adding that “the federal government's failure to speculate is a priority.”
A government spokesman said: “We are committed to improving the UK’s supercomputing capabilities.”
They pointed to the government-funded Isambard AI project on the University of Bristol, which they are saying might be one of the powerful machines of its kind on the earth when it comes online next 12 months.
