Justin Sullivan | Getty Images
Developers in the world of artificial intelligence can’t get enough of it nvidia Treatments. Demand is so strong that the company said late Wednesday that revenue in the current quarter will jump 170% to about $16 billion.
Nvidia shares rose more than 2% on Thursday before falling at the end of the day to close flat and miss a record close, while the broader market had a rough day.
There is another side to the story. AMDNvidia’s main competitor in the graphics processing unit (GPUs) market, is falling further behind, while the chip giant Intel Still miss the hottest trend in tech.
AMD and Intel shares fell 7% and 4%, respectively, after Nvidia announced its second-quarter financial earnings.
Nvidia’s huge report and executives’ comments that demand will remain high over the next year give investors reason to wonder if the company has any serious competition when it comes to making the kind of GPUs needed to build and run large AI models.
Nvidia’s success also signals a shift in the data center chip market. The most important – and generally the most expensive – part of building a data center is no longer tied to the central processors or CPUs made by Intel or AMD. Rather, it’s the accelerated AI GPUs that the big cloud companies buy.
Alphabet, Amazon, Meta and Microsoft Nvidia is buying Nvidia’s next-generation processors, which is so profitable that the company’s adjusted gross margin increased 25.3 percentage points to 71.2% in the period.
“NVDA data center revenue is now expected to be more than double INTC + AMD data center revenue combined, underscoring the growing importance of accelerators to today’s data center customers,” Deutsche Bank analyst Ross Seymour wrote in a note Thursday.
Nvidia is now expected to record $12 billion in data center sales in the current quarter, according to FactSet data. Intel’s data center group is expected to generate $4 billion in revenue, while analysts expect the AMD division to generate $1.64 billion in sales.
AMD and Intel are trying to stay relevant in the AI market, but it’s a struggle.
Intel CEO Pat Gelsinger said on the chipmaker’s earnings call in July that the company still saw “continued weakness” across all of its business segments through the end of the year and that cloud companies were more focused on securing graphics processors for AI rather than processors. Intel mainframe. Intel’s next high-end data center GPU, called Falcon Shores, is expected to be released in 2025. The 2023 tranche has been cancelled.
AMD said Thursday that it has acquired a French artificial intelligence software company called mysology. The company is also working on its own software suite for AI developers called ROCm to compete with Nvidia’s CUDA offering.
As with Intel, AMD faces a timing challenge. Earlier this year, it announced its new flagship AI chip, MI300. But they are currently only shipped in small quantities, a process called “sampling”. The chip is scheduled to hit the market next year.
“There is no meaningful competition for Nvidia’s high-performance GPUs until AMD begins mass shipping new AI accelerators in early 2024,” Raj Joshi, senior vice president at Moody’s Investors Services, said in an email.
The window closes. As AMD and Intel develop AI technology, they may find that all of their large potential customers are satiated with Nvidia chips before they can start shipping large quantities.
“AI spending will be a key driver for many of the companies in our coverage,” Joseph Moore, an analyst at Morgan Stanley, wrote in a report. Moore cited AMD, Marvel and Intel as “strong AI prospects”.
“But for those companies, the power of AI will be more than offset by crowding out the budget,” he wrote.
He watches: get rid of nvidia?