• The generative AI (GenAI) tech specialist reports a 265% year-on-year rise in fourth-quarter revenues
• The hyperscalers can’t get enough of its hardware and software
• And it’s still the only one with the ‘picks and shovels’ for the GenAI era, according to one seasoned analyst

Nvidia, the company at the heart of the generative AI (GenAI) revolution, is still in a league of its own in terms of financial performance and its unique ability to deliver the hardware and software components needed for generative AI training and inference. That’s the main takeaway from the company’s fiscal fourth-quarter results, which included a 265% year-on-year increase in revenues to $22.1bn and an almost laughable 983% year-on-year increase in operating profit to $13.6bn, better than expected. 

Of those fourth-quarter revenues, $18.4bn (about 83%) came from its datacentre products that are in significant demand from the likes of Amazon Web Services (AWS), Google, Meta (Facebook), Microsoft and more.  

For the full fiscal year that ended 28 January 2024, Nvidia reported revenues of $60.9bn, up by 126% compared with the previous financial year, while its fiscal year operating profit came in at almost $33bn, up by 681%.

Not only that, Nvidia’s management team expects revenues for the first quarter of the new financial year (fiscal 2025) to be about $24bn. You can get the full details, including a breakdown of the revenue lines, in this press release. 

Ahead of the financial report’s publication, there appeared to be concerns that Nvidia’s numbers would not live up to expectations and the company’s share price had been heading south during the first few days of this week to end Wednesday trading at $674.72. But investors jumped back on board once the numbers were public, with Nvidia’s share price climbing by almost 14% in after-hours trading to $769.01, a stock price that gives the chip and software giant a market valuation of about $1.9tn. The numbers are staggering and it’s hard to imagine the company won’t become a $2tn company in 2024.

The size and market power of Nvidia in a world that is clamouring to benefit from the GenAI applications that, currently, need Nvidia technology for optimal performance, is certain to be a big talking point at the upcoming MWC24, as we noted previously – see Nvidia is the one to watch at MWC24.

And the company isn’t sitting on its laurels: It has teamed up with Google to develop “optimisations” across its AI technology for Google’s new and relatively lightweight language models, dubbed Gemma, that can run “anywhere” (which means they are small enough to run on laptops as well as in the cloud). According to Nvidia, there are 100 million high-performance PCs running its RTYX GPUs (graphics processing units) around the world that could run Gemma and, thus, can be targeted by developers using the new model. See this announcement for more details.   

So, currently, Nvidia is king of the AI hill, but how long before its crown loses its lustre or even gets knocked off? Even though a number of its current main customers – the hyperscale datacentre operators – are developing their own AI-optimised chips and the networking technology that connects them, and rivals such as AMD and Intel are breaking sweat to catch up, Nvidia is still untouchable by any rivals, according to experienced technology and investment analyst Richard Windsor. 

“The ever-increasing number of competitors both in the merchant market and in-house remain unable to make a dent in Nvidia’s proposition,” noted Windsor in his latest Radio Free Mobile blog. “This is a result of its Cuda development platform, which is far superior to anyone else’s and has become an industry standard, and its breakneck product cadence, which ensures it always has the most advanced products in the market,” he added.

“The reality is that Nvidia has more demand than it can deal with, and so its revenues for the next few quarters will be determined by how much capacity it has purchased from TSMC [Taiwan Semiconductor Manufacturing Company] rather than how much customers have ordered,” added Windsor, who notes that this gives Nvidia good visibility about how much it has to sell and what its revenues will likely be in the near future. 

And the trends are looking good for the coming year, during which the focus of GenAI processing will shift from training large language models (LLMs) – the process of teaching AI models to perform a specific task or set of tasks by exposing it to massive amounts of data and which has been the dominant workload up to now – and more towards inference, the running of AI applications or workloads after models have been trained. 

“As 2024 is going to be the year of inference and as almost all of it is not ready to move out from the datacentre,” where Nvidia’s technology dominates genAI workloads, “Nvidia will continue to benefit… one thing is clear from the results and [earnings] conference call, which is that Nvidia is the picks and shovels of generative AI and, try as they might, no one else is currently getting a look in… It was the tools companies that made money out of the gold rush in the USA and the diamond rush in South Africa, and for this rush, Nvidia is the only tool company in town,” concluded Windsor. 

Now the big question is – what is Nvidia going to do with all of that cash? It has been making some investments in other companies, but nothing major, and buying back shares, but it must be considering its options for the time when other companies have managed to develop their own GenAI picks and shovels. 

In the meantime, Nvidia is basking in the corporate opportunity of a lifetime and making the most of it – now let’s see how its evolving strategy for the telecom sector plays out. 

- Ray Le Maistre, Editorial Director, TelecomTV