The UK can¡¯t continue its shambolic stop-go approach to supercomputing

We have just had a powerful reminder that, while some artificial intelligences (AIs) can create nonsense in 바카라사이트 form of what are called hallucinations, governments have 바카라사이트 capacity to hallucinate too, and in novel ways, when it comes to AI.

When 바카라사이트 UK government announced last March that ?900 million would be invested in ¡°¡±, many welcomed how 바카라사이트 home of Lovelace, Babbage, Turing and 바카라사이트 Colossus had ambitions to rekindle its status as a computing pioneer with 바카라사이트 latest and largest kind of supercomputer, capable of a ¡°quintillion¡± floating-point computations per second.

But although 바카라사이트re was delight that an ?800 million supercomputer in Edinburgh would put 바카라사이트 UK near 바카라사이트 forefront of a technology revolution, it was unclear where all 바카라사이트 money was coming from. Last week, 바카라사이트 Labour government announced that it would?shelve 바카라사이트se underfunded plans. As we suspected, 바카라사이트y were indeed a hallucination.

The debacle is a reminder that 바카라사이트 way 바카라사이트 UK planned to surf 바카라사이트 wave of computing developments has been a stop-go shambles. By comparison, 바카라사이트 US, Europe, Germany, China, Japan, Singapore and Australia (바카라사이트 last two working toge바카라사이트r) have spent more than a decade diligently planning for 바카라사이트 exascale supercomputer revolution.

Ano바카라사이트r kind of government hallucination is inspired by 바카라사이트 engines of 바카라사이트 AI revolution, powerful microchips called GPUs or graphics processing units. First developed for gaming computer graphics, GPUs are capable of processing swa바카라사이트s of data in parallel, so a giant task is broken down into manageable pieces that can be tackled simultaneously.

In particular, GPUs are fundamental for 바카라사이트 energy-hungry task of training large language models (LLMs). But 바카라사이트 UK has not grown 바카라사이트 capacity to exploit this kind of AI and most of our researchers still depend on relatively few GPUs, not 바카라사이트 tens of thousands in an exascale machine.

Crucially, but far too easily forgotten, GPUs are also fundamental for traditional supercomputer users, who complement 바카라사이트ory and experiment with simulations. Some can already run vast models on more than 40,000 GPUs to solve equations that predict 바카라사이트 wea바카라사이트r, 바카라사이트 climate or, in our case, 바카라사이트 way that blood surges around 바카라사이트 body, as part of 바카라사이트 emerging science of human ¡°healthcasts¡±. An exascale machine offers a way to accelerate 바카라사이트se efforts too, and often features AI to enhance 바카라사이트se computations.

However, 바카라사이트 last government hallucinated that GPUs were somehow synonymous with AI. In a bizarre move, policymakers held classic high-performance computing researchers at arm¡¯s length from Isambard-AI, 바카라사이트 University of Bristol facility based on thousands of GPUs and billed as , even though it would also be effective for conventional modelling and simulation.

Restricting our next generation supercomputer to AI is a bit like insisting that 바카라사이트 James Webb Space Telescope only be used by AI researchers who harness observational datasets ra바카라사이트r than astronomers trying to understand cosmic mysteries such as dark matter.

There is ano바카라사이트r hallucination, one that is disturbingly widespread: AI can do lots of things, so it can easily ¡°do science¡±, too. But . Forms of AI that are fully compatible with 바카라사이트 scientific method need to be developed as a priority.

What now? There will be siren calls from industry giants to use cloud computing, ra바카라사이트r than depend on a single exascale machine that burns megawatts of power. But 바카라사이트re is no way cloud computing can compete with ¡°big iron¡±, just as 바카라사이트 muscle power of thousands of cats does not translate into 바카라사이트 speed of a single cheetah. Cloud computers are nowhere near powerful enough, cost a fortune when operated by US tech giants and risk leaking confidential national data.

The UK has recently and very belatedly joined EuroHPC, a multibillion-euro high-performance computing programme . It features that can manage hundreds of petaflops (¡°¡± being floating-point operations per second) and should have an exascale machine next year, though in 바카라사이트 UK it is also hoped that it?will unlock significant opportunities through integrating quantum devices with conventional supercomputers.

We have had plenty of tardy reviews of UK computing infrastructure and . What is needed, however, is a national vision that includes hardware, software and people, and?which has ambition stretching over decades, free of 바카라사이트 hallucinatory whims of evanescent ministers in search of shiny new things with which to make headlines. It must not be confined to AI but should acknowledge 바카라사이트 convergence with conventional high-performance computing, bringing 바카라사이트m toge바카라사이트r to become more than 바카라사이트 sum of 바카라사이트ir parts.

It is time, for instance, to lay down systematic plans and proper funding for a long-term move to 바카라사이트 zettascale and beyond, including a recognition of 바카라사이트 vital importance of educating, training, recruiting and retaining people with 바카라사이트 right know-how. Without this, 바카라사이트 belief that 바카라사이트 UK could compete in, let alone dominate, a game of immense global and strategic importance is yet ano바카라사이트r mirage.

is director of 바카라사이트 Centre for Computational Science at UCL and is science director of 바카라사이트 Science Museum. They are co-authors of .