Nvidia is one of the clear beneficiaries of artificial intelligence—the company has been buying up its chips, and investor interest has sent its shares soaring to record highs. Nvidia’s sales surged last year as companies including Alphabet, Microsoft, Meta, Amazon and OpenAI purchased billions of dollars of Nvidia’s graphics processing units, the advanced components needed to develop and deploy artificial intelligence applications. and expensive chips. Many companies are currently in the building stage of artificial intelligence infrastructure. But will their millions or even billions of dollars of investment pay off? Clare Pleydell-Bouverie, a portfolio manager at Liontrust Asset Management, said on CNBC Pro Talks last week that companies that want to offer AI applications like ChatGPT to their clients need to build more infrastructure. This is because, she said, in order to implement artificial intelligence applications, companies must move from “general-purpose computing to accelerated computing.” “You can’t run artificial intelligence on traditional computing, it would be too expensive and too energy-intensive,” Priddle-Bouffry said. To meet this demand, an estimated 100,000 data centers will need to be built, she said. “This arms race to build artificial intelligence infrastructure has really begun,” she said, adding that Amazon, Google, Meta and Apple will spend $200 billion in capital expenditures this year alone. That’s a 35% increase from last year, she said, and all of this incremental investment will go toward artificial intelligence projects. “The natural resistance is – is this sensational investment worth it?” Pleidl-Buffry said. But, she said, “if you’re a cloud service provider, the ROI (return on investment) of accelerated computing is actually very compelling, so AWS (Amazon Web Services) or Microsoft are here.” “If you’re on Nvidia’s Spending $1 on accelerated computer architecture means $5 in immediate revenue over a four-year time horizon,” she said. These increases in capital spending will be funded by productivity gains, she said. Meta is one example, she said, with 50% of the content seen on Instagram being generated by artificial intelligence. “So you could actually think there’s going to be a pretty substantial return from this,” Priddle-Buffry said. She added that the world is “only in the first five minutes of building AI infrastructure.” However, she said some investments, such as Meta’s generative AI assistant and its Llama AI model, have “below-average returns” because they require “significant investment.” However, she said it will pay off in the long run. “When we think about the long-term goals, there’s probably no other company in Meta’s competitive landscape that has the ability to develop these artificial intelligence clusters,” she said. Meta’s Llama 3 model was built on a computer cluster using Nvidia’s 24,576 GPUs, which allowed Meta to achieve “an extraordinary breakthrough,” Pleydell-Bouverie said. “So we’re already starting to see the fruits of the labor of these tech giants,” she said. “However, there are many other companies that benefit from direct improvements in productivity as a result of these cloud service providers providing access to artificial intelligence.” She said companies using artificial intelligence to improve productivity include companies such as JPMorgan Chase using artificial intelligence cash management Bank of tools, and consumer beauty brands like L’Oreal that are entering the beauty tech space. Pleydell-Bouverie co-manages Lionstrust’s Global Technology, Innovation and Dividend Fund. All three funds outperformed their benchmark indexes in the year to March, with the Liontrust Global Technology Fund up 51.9%, higher than the MSCI World Information Technology Index’s 39.1% gain.
