Nvidia’s great silicon showdown

昔日盟友，渐行渐远

The relationship between Nvidia and the hyperscalers—cloud giants such as Amazon, Google, Meta and Microsoft—used to be straightforward. Nvidia designed and supplied chips; the hyperscalers built data centres using them. For now, the two sides still need one another (see chart). Yet both are preparing for a future in which they lean on each other less.

云巨头的自研芯片野心

Such moves are prompted, in part, by unmistakable signs that Nvidia's core customers are drifting away. The hyperscalers are no longer content only to buy Nvidia’s chips and are spending billions on designing their own. Google has for years rented access to tensor processing units (tpus), specialised ai chips, through its cloud. Now it is selling tpu systems to other firms. Amazon puts the annualised revenue of its custom-chip business, mostly tied to ai, at $25bn. Andy Jassy, the company’s boss, reckons that makes it one of the world’s three biggest data-centre chip businesses. Microsoft and Meta have also developed chips. Anthropic and Openai, two big ai labs, plan to do the same.

Hyperscalers have good reasons to design their own silicon. Chips account for much of the cost of an ai data centre. Bernstein, a broker, estimates that in a server rack running Nvidia’s h100 chips, priced at $25,000 apiece, spending on chips makes up three-quarters of the total cost. Custom silicon is a fifth to a third as expensive, though less powerful. The cloud giants argue that they still get more computing power per dollar. Custom chips are also better suited to particular jobs: Google’s tpus for calculations underpinning its ai models, for example, and Meta’s processors for recommendation algorithms.