LIn September, Nvidia, the US manufacturer of graphics processing chips, and the Japanese company SoftBank announced an agreement whereby Nvidia would acquire the British chip designer Arm from SoftBank for $ 40 billion. Given that SoftBank had acquired Arm in 2016 for $ 32 billion, it could be said that a 25% gain on a five-year investment should not be looked down upon, especially if industry murmurs About SoftBank’s quirky investment strategy and Arm’s internal difficulties with its China-based operation are to be believed.
But even if one were fool enough to sympathize with SoftBank’s desire to get out of the hole it had dug for itself, the idea that Arm should sell itself to a US chip. maker It’s so silly that even the Boris Johnson administration had started to smell a rat. And so Monday announced that the secretary of state for digital, culture, media and sport was “intervening in the sale on national security grounds”, based on advice received “from officials throughout the investment security community.” To which decision the only possible answer is: why did it take so long?
To fully appreciate the idiocy of allowing the Nvidia deal to go ahead, some technical background may be helpful. Arm is a notable Cambridge-based company that grew out of the early UK leadership in microcomputers in the 1980s, fueled by the BBC’s digital literacy project and the emergence of BBC Micro. This famous machine was designed by Acorn, a Cambridge startup, and provided the platform on which a generation of game programmers and designers first learned the basics of their craft.
After the success of the BBC machine, the question for the Acorn founders was: what’s next? His response was to design a new type of processor called a Risc (reduced instruction set computer) machine that had the potential for faster, cheaper, and more energy-efficient computing. The company they created to develop this idea was originally called Acorn Risc Machines, hence the acronym Arm, although it eventually became Advanced Risc Machines.
Arm was based on a clever idea from the start, but the real stroke of genius was its co-founders’ invention of a new business model. The company would not make chips, but simply design processors to exploit the Risc architecture and license the designs to anyone who wanted to make the chips while paying Arm royalties for the use of their intellectual property. This was clever because chipmaking is a viable capital-intensive business only for large companies with very deep pockets and an appetite for risk.
The key to Arm’s extraordinary growth was the smartphone revolution started by Apple’s iPhone in 2007. Suddenly, the world changed everyone who wanted a powerful handheld that could access the Internet. And all those little computers needed processors that were small, powerful, and most importantly, consumed as little power as possible. In other words, arm designs. The result: something like 95% of all mobile devices in the world now run on Arm-designed processors. That is a British success story on an unimaginable scale.
But here is the critical part: Arm does not manufacture chips, that business is left in the hands of so-called “silicon foundries”, that is, semiconductor manufacturing plants. If you need a processor for your new phone or tablet or other gadget and are willing to pay the royalties, you can use an ARM design. And so are your competitors. Arm can, and does, license its designs to all participants because is not involved in your business! You don’t compete with your customers.
Nvidia, however, does it make processors. Its graphics processing units (GPUs) are the wonders of the silicon world. Originally dominant in the areas of computer games, graphics processing, and movie animation, it turned out that they are also excellent for building machine learning systems and neural networks – everything that happens on desktops or in large racks of servers in communication centers. data. And as the frenzy for machine learning has turned frantic, Nvidia is on the rise, because it’s riding a tsunami.
But it doesn’t really have a serious foothold in the world of mobile computing, which is why it got hooked when SoftBank decided to ditch Arm. In the crazed world of corporate mergers and acquisitions, it probably sounds like a smart buy. But in the long run it is a way to cripple Arm (or maybe even destroy it) because every chip manufacturers they are ultimately competitors to Nvidia. They are in the same business and only the paranoid survive in that uproar.
If the Johnson administration is really serious about the UK continuing to have a real influence on the global digital economy, then letting the only company in the country that truly outperforms the world go to a US company) would be a certifiable insanity. And the annoying thing is that there is an obvious solution: The government should take a 51% stake in Arm to protect it from corporate raiders and let it continue what it does best.
What i have been reading
Free for all
TO sharp piece by James Ball in the New european about the real scandal of the David Cameron case.
A thoughtful meditation by Drew Austin on his kneeling bus blog about what the internet is doing to our sense of place.
George is Digismak’s reported cum editor with 13 years of experience in Journalism