The UK government will invest up to £100 million to procure thousands of high-powered AI chips. The procurement is part of Prime Minister Rishi Sunak's ambition to establish the UK as a global leader in AI technology and innovation.
Government representatives are reportedly in talks with leading chipmakers NVIDIA, AMD, and Intel to procure equipment to establish a national "AI Research Resource." This is seen as an integral component of Sunak's vision for the UK to ascend to a dominant position on the global AI stage.
While negotiations continue with various chip manufacturers, NVIDIA's dominance in producing AI-optimized chips make them the preferred choice. Their chips are currently used for training and running complex AI systems like large language models such as ChatGPT. The UK Research and Innovation agency is currently leading discussions on the acquisition of nearly 5,000 graphics processing units (GPUs) from NVIDIA.
For context, the most recent iteration of ChatGPT required training utilizing an astounding 25,000 NVIDIA GPUs. While the UK has exhibited aspirations of becoming an AI juggernaut, the reality reveals a significant lag behind counterparts in the US and Asia, especially concerning the requisite computing resources essential for the development and deployment of intricate AI models.
Earlier this year, a governmental review criticized the lack of dedicated AI compute resources, with fewer than 1,000 high-end GPUs available to researchers. It recommended making at least 3,000 top-tier GPUs accessible as soon as possible.
The new chips will provide the backbone for the national AI Research Resource to be launched in the UK next summer. It will support UK researchers and companies in innovating AI applications and maintaining international competitiveness.
This hardware investment aligns with the UK's national AI strategy outlined by PM Sunak. He aims to position the UK as a global hub for AI standards development and oversight. As part of his government’s drive, the UK is set to host the first "global" summit on artificial intelligence later this year. Stakeholders hope that this summit will pave the way for constructive collaborations between global governments and leading AI corporations, fostering the harmonious development of this transformative technology.
However, a major question looms as to whether or not China should be invited. Including Beijing risks upsetting allies and potentially limiting open information sharing at the event. But excluding the world's second largest AI power would significantly undermine the summit's international credibility. This comes amid escalating geopolitical tensions, as governments recognize AI technology will be key to future economic strength and national security.
Earlier this month, the Biden administration banned U.S. investments in three sensitive Chinese technology sectors: semiconductors and microelectronics, quantum information technologies and certain AI systems. This follows China's declaration in May that chips from US manufacturer, Micron technology posed national security risks and banned their use in critical infrastructure projects.
In the backdrop, other nations and global tech giants are fervently trying to secure these invaluable chips. Last week, reports emerged that Saudi Arabia had procured at least 3,000 NVIDIA H100 processors, each valued at $40,000. The UAE has also secured access to thousands of NVIDIA chips and has already developed its own open-source large language model, known as Falcon. Furthermore, companies like Microsoft, Amazon, and Google are also vying for a substantial share of these sought-after chips.
With the UK aiming to be an AI superpower, experts say much larger investments will be needed to seriously compete in the global microchip industry that is foundational for advanced AI systems.
The UK's current investment plans lag significantly behind other major powers in the AI chip race. In May, Sunak's government announced plans to invest £1 billion over 10 years in semiconductor research, design and production. But this pales in comparison to efforts by the US and EU. In 2022, the US passed the CHIPS and Science Act which included $52 billion to boost domestic chip production, and just last month, the EU enacted their €43 Billion Chips Act.
The UK's relatively modest investments in AI chips thus far suggest it faces an uphill battle to truly compete in the high-stakes global AI race. The current £100 million procurement for the new national AI Research Resource is unlikely to be sufficient without further multibillion-pound funding commitments.