Seen as the darling of the artificial intelligence (AI) wave, Nvidia’s latest quarterly earnings of US$22.1 billion will have been warmly welcomed by a number of Australian fund managers with strategic holdings in the chipmaker.
On Thursday, the company reported earnings grew 22 per cent from the previous quarter and were up 265 per cent from a year ago.
Nvidia’s founder and chief executive, Jensen Huang, attributed the monumental growth to surging global demand, which has propelled generative AI and accelerated computing to “the tipping point”.
“Our data centre platform is powered by increasingly diverse drivers – demand for data processing, training and inference from large cloud-service providers and GPU-specialised ones, as well as from enterprise software and consumer internet companies. Vertical industries – led by auto, financial services and healthcare – are now at a multibillion-dollar level,” Mr Huang said.
Looking ahead, Nvidia has outlined optimism for the technology, forecasting revenue to hit $24 billion, plus or minus 2 per cent, for the first quarter of fiscal 2025.
In Australia, a number of funds deliver exposure to this underlying technology among their top 10 holdings, including Betashares’ Global Sustainability Leaders ETF and Betashares Nasdaq 100 ETF; the Alphinity Global Equity Fund; the ARK Global Disruptive Innovation Fund, distributed by Yarra Capital Management; and the Munro Global Growth Fund.
Kieran Moore, portfolio manager at Munro Partners, observed that Nvidia’s data centre division grew over 400 per cent compared to the fourth quarter last year, and is seeing accelerating demand for its high-end semiconductor products that enable corporates such as Microsoft, Alphabet, Meta, Amazon and many others to deploy artificial intelligence applications.
“For the next quarter, the company guided to revenue growth of over 200 per cent year-on-year again, showing the continued strength in demand for their products,” Mr Moore told InvestorDaily.
Last month, Munro had pegged Nvidia to lead the charge in a tech rally in 2024, and said it could go on to overtake Apple to become the next tech Goliath.
“At Munro Partners, what we look for is companies that have the ability to double their earnings over the next five years. For Nvidia, despite the strong performance, we believe the stock can still double its earnings over a five-year period because those earnings are backed by the structural tailwind that is AI,” he added.
Presently, the fund manager holds the belief that less than 20 per cent of data centres have adopted accelerated computing technology, which will be needed to power the AI applications of the future.
“In order for this penetration to play out, Nvidia plans to continue to release new products that enable this structural shift to occur, and we expect to hear further information throughout the year as these products come to market,” Mr Moore said.
According to Betashares senior investment strategist Cameron Gleeson, Nvidia had managed to successfully defy expectations again after it beat analyst expectations by anywhere between 10 and 20 per cent.
“The market provides a bullish estimate on earnings, and Nvidia responds by exceeding expectations with an impressive beat,” he said.
“It’s pretty easy to see how Nvidia has become the poster child of the Magnificent Seven rally with significant gains in revenue associated with its H100 chips. Companies around the world are scrambling to get their hands on these chips to power their own artificial intelligence ambitions.”
Since the results announcement, Nvidia’s share price witnessed growth of over 16 per cent throughout the day. Meanwhile, the S&P 500 saw its largest daily gain in 13 months and the Nasdaq Composite finished just shy of a record high.
Some investors remain on the fence
However, there are many in the market who remain wary of jumping on the Nvidia bandwagon although they are optimistic about the long-term potential of artificial intelligence.
Speaking to InvestorDaily, AMP’s chief economist and head of investment strategy, Shane Oliver, noted that the AI thematic is supported presently by massive revenue growth for Nvidia, unlike for the tech stocks in 1999.
“AI has a strong long-term future. But history tells us that good things can be pushed to extremes by ebullient investors,” he said.
“So, investors need to tread carefully to make sure they don’t get sucked in at the wrong time.”
Similarly, Platinum Asset Management portfolio manager Jimmy Su has previously voiced caution that current spending in the theme was “funded by private markets and not by adoption among consumers and businesses”.
In a quarterly update for the Platinum International Technology Fund at the end of December 2023, he said investor expectations imply a level of capex spending which may prove to be too high.
He outlined concerns regarding the vulnerability of short-term demand and the risk that investors’ expectations for Nvidia over the next three years are too high.
Despite popular belief, generative AI adoption and demand for graphics processing unit (GPU) capacity among consumers and enterprises remains minimal, Mr Su said.
“The majority of demand is driven by private capital which has flooded the sector. A portion funded AI application companies who are renting thousands of GPUs from large cloud service providers (the big three being Amazon AWS, Microsoft Azure, and Google Cloud) to train new AI models and build AI applications,” he explained.
“This, in turn, is causing a capacity shortage at the cloud service providers, who, in turn, are scrambling to buy more GPUs. The remaining portion of capital-funded small cloud service providers who are buying up as many GPUs as they can and renting them out as cheaply as they can. Neither of these groups are profitable or have sustainable business models (yet) and it’s fair to assume that future GPU demand will be highly dependent on the ability of these companies to raise money and the willingness of investors to fund them.”
Additionally, consensus estimates forecast Nvidia to generate around $170 billion in revenue from generative AI and given that GPUs make up some 70 per cent of data centre capex for AI workloads, this would imply that cloud service providers (CSPs) will invest approximately $240 billion in capex over the next three years, Mr Su said.
“In our view, it’s unlikely that the CSPs are willing to make a risky bet of this magnitude in such a short time frame on such an immature technology. We have calculated where we think AI adoption could be in three years’ time and think the CSP industry will make ~$45 billion in revenues from generative AI. Such level of revenues on ~$240 billion of capex suggest capacity utilisation of less than 30 per cent and industry losses of ~$25 billion per year,” he explained.
