Meta Platforms is planning to pay Nvidia billions of dollars this year for its cutting-edge AI technology. But the chipmaker isn’t the only winner in this massive transaction. In an Instagram video Thursday, Meta CEO Mark Zuckerberg said the social media giant is on track to own 350,000 Nvidia H100s by the end of the year, as it advances its AI efforts. Including other AI-focused chips, Meta’s compute infrastructure will be equal to almost 600,000 H100 equivalents, Zuckerberg said. Meta has been a significant Nvidia customer for years, but Zuckerberg’s comments Thursday highlight the magnitude of demand for its pricey graphics cards, or graphics processing units (GPUs). The remarks also add to the bullish AI picture painted this week by chipmaking giant Taiwan Semiconductor Manufacturing Company , which produces Nvidia’s processors. “This is just a very exciting time to be Nvidia, because they’ve got Taiwan Semi working around the clock to make as many chips as they need, Jim Cramer said Friday. Nvidia’s stock has been off to a blistering start this year, up more than 20%, including a 3% gain Friday. Nvidia is on pace to set another all-time high, at $589 per share. Shares of Meta added nearly 2% Friday, to roughly $383 each, putting the stock on pace for a new all-time high for the first time in more than two years. Its prior record close came on Sept. 7, 2021, at $382.18 per share. The tech research firm Omdia estimated Meta bought about 150,000 H100s in 2023, analysts at Wells Fargo said in a note to clients Thursday. Against that backdrop, Meta’s plans to end 2024 with more than double that estimate is encouraging for Nvidia, Wells Fargo argued. “We see this as continued positive validation that the AI infrastructure buildout remains in its early innings,” the analysts led by Aaron Rakers wrote. The H100 is Nvidia’s top-of-the-line AI chip. It was in short supply throughout 2023 amid a spending boom in generative AI technology sparked by the launched of viral chatbot ChatGPT in November 2022 . Nvidia began shipping the H100 to customers in late 2022, with deliveries picking up throughout last year. The H100 supplanted its less-powerful predecessor, called the A100, as the most-coveted chip to use for training AI models, a compute-intensive process in which computers look for patterns in large quantities of data. GPUs are well-equipped to handle this task due to their ability to perform many calculations at once, known as parallel processing. Estimates on Nvidia’s share of the AI training market vary, but is generally thought to be well above 80%. NVDA 5Y mountain Nvidia’s stock performance over the past 5 years. Analysts expect Nvidia’s revenue in its fiscal 2024 — which ends this month — to total roughly $59 billion, representing a 119% increase from the prior 12 months, according to FactSet. Nvidia’s adjusted earnings per share are projected to nearly quadruple over the same period, to $12.30 from $3.24, FactSet data shows. Analysts on Wall Street are modeling for more strong growth in Nvidia’s upcoming fiscal 2025 which begins in less than two weeks, with revenue expected to increase around 58%, to $93.2 billion, and adjusted earnings increasing 71%, to $21.04 per share. But the pace of growth is expected to slow considerably in Nvidia’s fiscal 2026 — a 12-month period that primarily takes place during calendar 2025. Indeed, investor concerns about whether 2025 will be a so-called “digestion” year for Nvidia, with new orders for its chips slowing down, contributed to the stock’s relatively sideways trading in the final few months of 2023. Still, Nvidia’s stock more than tripled in 2023, leading the S & P 500, and has so far been a big winner in 2024. It remains one of two stocks that Jim has bestowed as an “own it, don’t trade it” position; iPhone maker Apple is the other. While acknowledging that Nvidia’s growth will moderate over time, the company has proven an ability to innovate and deliver more powerful chips, and these new product cycles should help stoke demand among customers. Plus, its push into software and service offerings should help to smooth out the cyclicality in its business, given hardware has historically been prone to boom-and-bust cycles. META 5Y mountain Meta Platforms’ stock over the past 5 years. Zuckerberg’s update on Meta’s growing war chest of AI chips and AI initiatives more broadly help support our continued investment in the stock, which was our second-best performer last year behind Nvidia. A key prong to our thesis is its AI efforts, which to date have improved ad targeting on Instagram and Facebook and boosted user engagement with Reels, its short-form video feature that competes with TikTok. While investors — including us — want to see Meta’s big bet on its money-losing Reality Labs division start to pay off , spending money on AI chips that also enhance profitable parts of its business is likely to be taken in stride by the market. Elsewhere in the portfolio, Meta’s AI chip plans should help Eaton Corp , which makes electrical components, like circuit breakers, and power systems and has significant exposure to the fast-growing data center industry. As Meta secures more AI chips, they will be placed in data centers where their computational capabilities will be utilized. However, data centers with a lot of GPUs for AI tasks require a lot more electricity than facilities designed to run general-purpose workloads, which usually contain a lot of central processing units, or CPUs. The additional power requirements of AI-focused data centers benefits Eaton, which we added to the portfolio in November . On the company’s third-quarter earnings call in late October, management said the impact of AI data center buildouts is “just starting to show up in our order book.” We’ve been hoping for a pullback in Eaton’s stock price to buy additional shares, but the stock is performing well, up more than 6% since Nov. 15 when we took a stake. (Jim Cramer’s Charitable Trust is long META and NVDA . See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Meta Platforms is planning to pay Nvidia billions of dollars this year for its cutting-edge AI technology. But the chipmaker isn’t the only winner in this massive transaction.
Read the full article here
