Late last week technology firm Nvidia gave the listed tech space a much-needed shot of optimism, one of such epic proportions that many are feeling the flush of bullishness returning in a big way and assigning the past year’s body blows to the stuff of distant memory.
This has been a story months in the making. In May, accounting for the first quarter of 2023, Nvidia shares demonstrated a whopping 36% swell, according to S&P Global Market Intelligence. “The sharp jump has increased Nvidia’s year-to-date gain to 164.3% even as the chipmaker’s market capitalisation approaches the $1-trillion mark,” a Motley Fool article said at the time. It has since danced in and out of that trillion-dollar club.
Beyond the actual cash-in-the-bank numbers achieved, the promise of demand to come set investor hearts aflutter, or as CNBC more stoically put it: “Its sales forecast of about $11bn for the second quarter — more than 50% higher than Wall Street estimates — was what sent the US giant’s stock surging more than 24% in after-hours trade”.
And last week, as mentioned, the good news kept coming for Nvidia, which reported on the second quarter of 2023 (ending July 30) late last Wednesday. It showed year-on-year revenue was up 101% and earnings per share 429%, blowing internal and external expectations out of the water.
The engine of this growth is Nvidia’s data centre business, including the A100 and H100 chips that enable generative artificial intelligence (AI) systems, such as Open AI’s ChatGPT. Aha — light dawns on marble head! If you hadn’t already known this about Nvidia, at least now the phenomenal growth figures of the prior six months will suddenly make perfect sense ... for anyone not living under the farthest-flung rock.
This is because the one tech story that has reliably eclipsed both depressive layoff news and the soap opera drama from the cultural front lines of Twitter-turned-X, is generative AI. For accuracy’s sake, I feel compelled to clarify that Nvidia designs and sells graphics processing units (GPUs) and chip systems, used in complex high-demand processing such as gaming, crypto mining, robotics and vehicles.
It doesn’t actually make its chips. Those come from Taiwan Semiconductor Manufacturing Company (TSMC) — a core nodule of the broader supply chain that is also reaping the benefits of Nvidia’s boom. When crypto mining underwent a mini-revolution in 2022 it seemed Nvidia would be left hurting as demand crashed, and stock prices too. And, “then it happened”, as Financial Mail columnist Duncan McLeod put it, “it” being the unforgettable, unpredictable, viral-doesn’t-cover-it explosion of interest in generative AI.
So generative AI is powering Nvidia and Nvidia is powering generative AI, with tremendous and sustained demand the icing on the cake.
McLeod did a wonderful write-up on this twist of Nvidia’s fate in June, writing: “All these companies need silicon to power their AI tools — truckloads of it, to fill giant cloud data centres. And one company happened to have just released the best AI chip money could buy: Nvidia.”
So generative AI is powering Nvidia and Nvidia is powering generative AI, with tremendous and sustained demand the icing on the cake. It’s not quite “the only game in town” as some have called it, but it is certainly the dominant and most valuable player. The lay of the chipland will change though, and soon. British chip designer Arm has begun the process of listing on the Nasdaq, in what is expected to be one of the biggest flotations of the year.
Arm registered this intent last Monday, having decided against a UK listing and off the back of a failed Nvidia buyout in 2021. Those big plans were scuppered by the UK competition regulators. In July, the Guardian and Financial Times reported that Nvidia was “in talks” to be an “anchor investor” for the initial public offering (IPO), though I don’t believe this has been confirmed anywhere yet.
Arm’s got much going for it, but is primarily associated — at least these days — with licensing its chips for low-power-consuming devices, such as wearables, phones and tablets, as well as Apple’s M1 and M2 chips, which are built on Arm architecture. However, a large injection of cash from an IPO could set it up nicely to start handling the gaping maw of AI chip demand.
Some commentators also swear by the speeds of AMD’s chips, suggesting their undervaluation. An AI software start-up called MosaicML made headlines in July with a report claiming that AMD’s M1250 outperforms Nvidia’s A100 by as much as 80%. Collectively, these and a smattering of others who make up the traditional chip players are well-positioned to ride this demand wave, but shuffling the existing deck is not the only foreseeable change.
In July, the Financial Times reported on a handful of chip start-ups “rushing to capitalise on the surging demand for the specialist chips”, calling the shortage of these a “once-in-a-generation opportunity for new challengers”. The FT name-dropped SambaNova, Graphcore and Tenstorrent in particular.
It is a space seeing huge investment right now, and with that will come the innovation, increased competition, and potentially the next huge leap, that we can’t even begin to articulate, something more earth-shattering than outsourcing your legal or accounting functions to AI.
Generative AI is changing the world every day, but not always in the ways that immediately spring to mind.
• Thompson Davy, a freelance journalist, is an impactAFRICA fellow and WanaData member.
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.