Information technology has two major parts: computing and communication. The next big step in computing is processors for data centres and AI. Both are critical components of an IT infrastructure. It is as important to our daily lives as water, transport and electricity supply.
Every nation has a sovereign duty to provide this infrastructure and protect it from foreign interference. As IT infrastructure relies heavily on technology, this is an issue of tech sovereignty.
To achieve such sovereignty, particularly for crucial components like processors, we have to ask: do we have it in Britain? If not, do we have supply chains with stable suppliers? Still if not, do we have access to monopoly or oligopoly suppliers? If Arm is to be sold to Nvidia, the answer to these questions will be no.
In 1985, my technical team at Acorn Computers implemented the world’s first commercial Reduced Instruction Set Computer (Risc), which we named the Acorn Risc Machine (Arm). In 1990 I persuaded Apple to help us spin out Arm from Acorn because I argued that processor customers would not want to buy a product from a competitor.
Arm has become the UK’s most successful technology company because of its neutral licensing model.
It has over 500 licensees globally, which ensures British companies have unfettered access to licences and the choice of Arm-based processor products from all over the world.
Arm is still a British company, although it was bought by SoftBank, a Japanese financial firm, in 2016. Arm’s growth has stalled because SoftBank, a financial giant in the tech world, insisted on it following an expansion into the internet of things (IOT), instead of designing a new generation of Arms for data centre servers.
Nvidia has made a $40bn (£30.8bn) bid to buy Arm to address the data centre server market – the core business of a failing Intel. Nvidia plans to take the microprocessor crown from Intel by winning the server market based on Arm architecture.
To understand the implications of such a takeover, we need to consider the markets for the three basic types of processor: scalar processors that power servers, PCs, smartphones and the IOT, and vector processors for graphics – not to be confused with graph processors for AI. The UK has good access to scalar processors with Arm for smartphones and IOT, but it suffers from the duopoly of Intel and AMD in PCs. Apple has just moved from Intel to Arm for its iMac PCs.
Arm has a chance to move into this space and increase competition for PC processors.
The graphics market is a duopoly at the high end with Nvidia and AMD, with suppliers like Arm, Imagination and others in smartphone graphics.
The real battleground of the future, though, is data centres using servers, and graph processors for AI. Intel has a dominant position, together with AMD, but has stumbled at the latest semiconductor nodes. This is the great opportunity for Arm to resume its growth by developing new server processors and licensing them around the world. This would increase market choice beyond Intel and AMD.
Last but not least, they can set a new standard of graph processors for AI. Fortunately, Britain has the leading company in this space, Graphcore, which has designed an architecture custom-made for AI graph problems.
Nvidia’s recent stratospheric rise is due to its high-performance graphics processor units (GPUs). Graphcore outperforms the best Nvidia GPU chips by a substantial margin.
I applaud Dominic Cummings’ ambition to create a £1 trillion firm in Britain. If there ever was the prospect of an industrial strategy, then starting with the dominant position of Arm in smartphones and IOT, and adding next-generation server products, must be the best chance we have to achieve it, especially if combined with Graphcore’s AI capabilities.
This would make Arm the most comprehensive supplier of processors to the world.
Such a strategy can happen by taking Arm public on the LSE. The Government would be the cornerstone investor with $1bn-$2bn and orchestrate investments from the main licensees of Arm, like Apple, Samsung, Sony, Infineon, NXP, Qualcomm and Nvidia. They all have an interest in Arm’s neutral status, rather than a Nvidia monopoly.
This can work for Britain on multiple fronts, electrifying the languishing tech stock sector of the LSE; retaining choice of all processor types for the UK and the world; and preventing the US controlling all the processor architectures globally.
Britain’s processors must not become collateral damage in the trade war between the US and China. Let us have an ambitious industrial strategy and build that first £1 trillion company.
Hermann Hauser is co-founder of Arm and Amadeus Capital