UK Prime Minister Rishi Sunak has approved the allocation of £100 million of public funds for the procurement of semiconductors to bolster the country’s AI sector and enhance computing resources. The government is reportedly in discussions with major US microchip companies, including Nvidia, AMD, and Intel, to secure chips for AI technology development.
The procurement process will be overseen by UK Research and Innovation (UKRI), which is reportedly close to finalizing an order for 5,000 graphics processing units (GPUs) from Nvidia. Nvidia’s GPUs play a crucial role in powering generative AI models like OpenAI’s ChatGPT. However, concerns have been raised that the order may be insufficient, as the latest version of ChatGPT was trained on around 25,000 Nvidia chips.
Civil servants are reportedly advocating for additional funds to be allocated for AI infrastructure projects, beyond the current procurement effort. The Department for Science, Innovation and Technology (DSIT) mentioned that the additional funding through UKRI would complement a separate £100 million investment to establish the Foundation Model Taskforce.
The announcement of the funding has prompted inquiries from ministers about the distribution of AI funding. Some have sought clarity on how the Foundation Model Taskforce plans to spend its funding. The funding initiative aligns with Sunak’s ambition to position the UK as a science and technology superpower, with a strong focus on AI.
Sunak has previously approved £100 million for a generative AI taskforce and a £1 billion semiconductor strategy. However, there have been criticisms that the level of support remains inadequate compared to international competitors. The US and EU have allocated significantly larger sums for semiconductor development, at $50 billion and €43 billion respectively, in comparison to the UK’s allocation.
The UK is also scheduled to host a global summit on AI safety in November, though there are uncertainties about the event, such as whether China will be invited.