Advanced Micro Devices has jacked up its market forecast for artificial intelligence chips to an eye-popping $400 billion by 2027. The chipmaker is sure to see its slice of that pie grow in the coming years, but the industry’s dominant player, Nvidia , is well-positioned to handle the emergent competition. Just a few months ago, AMD CEO Lisa Su projected the AI chip market would be worth $150 billion by 2027 amid accelerating investment in the technology since OpenAI’s viral chatbot ChatGPT launched in late 2022. Alongside officially launching its MI300X chip to challenge Nvidia, AMD on Wednesday also revised that forecast up to $400 billion from $45 billion this year, representing a compound annual growth rate north of 70%. The complex language model behind ChatGPT was trained using Nvidia’s chips . Overall, Nvidia commands about 90% of the market for AI training, Bank of America analysts estimate. It’s impossible to say whether AMD’s market-size prediction will materialize. Indeed, some Wall Street analysts — including those at Citigroup, Bernstein and Morgan Stanley — said it seems overly optimistic. The skeptics may be proven right. But at this point, the trajectory of AMD’s AI chip forecast matters more than the specific number. The company believes significant increases in computing power will be needed to meet AI workload demands down the road. It’s more about market growth than its market share. In an AI world as large as AMD expects, multiple species of artificial intelligence chips can thrive. “I think there’s room for both, clearly, because there’s just so much demand,” Jim Cramer said Thursday. “But the software layer of Nvidia still gives them the edge. Not the actual hardware. They seem almost even.” AMD shares surged more than 9% on Thursday and have nearly doubled year to date. Club stock Nvidia, which has more than tripled this year, added 2%. AMD, a former Club holding, joined our Bullpen watchlist of stocks in early October, due to our belief the company can be a No. 2 player behind Nvidia in the AI chip market. NVDA AMD YTD mountain Nvidia’s year-to-date stock performance versus AMD. Nvidia’s combination of hardware and software has been the secret sauce behind its AI leadership to date. In 2006, Nvidia launched a propriety software called CUDA that allowed developers to tap into maximize the computing capabilities of its chips. CUDA is still an important part of Nvidia’s competitive position, even as AMD makes progress on its open-source answer known as ROCm. More recently, Nvidia’s attractiveness as an investment has been bolstered by a different software opportunity. With services such as DGX Cloud , a supercomputer accessible via web browser, and AI Enterprise, a suite of tools for companies to develop their own artificial intelligence applications, Nvidia has a burgeoning software-and-services revenue stream that, as of now, AMD does not. The MI300X processor appears to be a formidable piece of hardware compared with Nvidia’s H100, which launched earlier this year. The H100 has been in short supply amid an influx of orders from cloud service providers and other tech giants. Club holdings Microsoft and Meta Platforms , as well as OpenAI, are among the companies placing orders for AMD’s chip — evidence there’s an appetite for both more computing power and alternatives to Nvidia. Cloud providers, in particular, stand to gain from AMD adding to the availability of high-quality AI chips as they try to get their hands on as many as possible to power a growing number of AI applications. AMD has said it expects AI chip revenue to be at least $2 billion in 2024 , which would be roughly 7.5% of companywide revenue, according to analyst estimates compiled by FactSet. By comparison, Nvidia’s data-center division — home to sales of its AI chips — is expected to generate more than $77 billion in revenue in the 12 months ended January 2025, per FactSet. While Nvidia’s networking business also goes into data-center sales, it’s clear its AI chip business dwarfs AMD at this stage. That’s also reflected in their market capitalizations: $1.14 trillion for Nvidia and $206 billion for AMD. The MI300X and H100 offer roughly comparable performance in training large AI models, AMD contended, while claiming its chip offers improvements over the H100 in a process called inference. That’s basically the day-to-day deployment of an AI model after it’s been trained on a ton of data. To be sure, Nvidia is set to launch next year an updated version of the H100 — branded as the H200 — that is better suited for inference than its predecessor . That should enable Nvidia to “regain the performance lead,” Citigroup said in a note to clients Thursday. Analysts also expect Nvidia to announce an entirely new AI chip at some point in 2024 called the B100. At this point, AMD is fully in the AI chip race alongside Nvidia. There can be more than one winner. (Jim Cramer’s Charitable Trust is long NVDA, MSFT, META. 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.
Nvidia CEO Jensen Huang speaks at the Supermicro keynote presentation during the Computex conference in Taipei on June 1, 2023.
Advanced Micro Devices has jacked up its market forecast for artificial intelligence chips to an eye-popping $400 billion by 2027. The chipmaker is sure to see its slice of that pie grow in the coming years, but the industry’s dominant player, Nvidia, is well-positioned to handle the emergent competition.
Denial of responsibility! Secular Times is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – seculartimes.com. The content will be deleted within 24 hours.
