AMD fires latest shot at Nvidia in AI-chip war

Muhammad Ali versus Joe Frazier. Jack Johnson versus James Jefferies. Rocky Marciano versus Jersey Joe Walcott.

Related: Why a veteran fund manager sold some Nvidia stock before its stock split

Fight fans will argue over their favorite heavyweight bouts, but it’s pretty much agreed that when the big guys get into the ring, the match is going to get pretty exciting.

The exploding artificial-intelligence sector has its own breed of heavy hitters vying for the championship.

And make no mistake, there’s a lot of money in this fight club.

The value of the generative-AI market is poised to explode, growing to $1.3 trillion over the next 10 years, according to a report from Bloomberg Intelligence. AI-related hardware could reach $640 billion by 2032 from less than $40 billion in 2022.

Generative AI, the artificial-intelligence production of original text, images, videos and other data, could expand to 10% to 12% of total information-technology hardware, software, services, advertising and gaming spending by 2032 from less than 1% today, Bloomberg estimates.

AMD appearance at trade show

AI-related hardware could reach $640 billion by 2032 from less than $40 billion in 2022 and the report said that Microsoft (MSFT) , Apple (AAPL) , Alphabet (GOOG) , Nvidia (NVDA) and Amazon (AMZN) are most exposed to the opportunity.

Nvidia is the current AI-chip champion, having recently crossed the $3 trillion market-capitalization threshold and surpassing Apple to become the world’s second most valuable company, smack dab behind Microsoft.

Related: Analysts race to reset HPE stock price targets as AI powers earnings

Intel (INTC) is also looking for its piece of AI action. On Tuesday, the company unveiled Xenon, a data-center processor that Intel said would require fewer racks and less power while delivering better performance.

The Santa Clara, Calif., company also launched its Lunar Lake chip for AI computing, which it says will start shipping in the third quarter, as it looks to build on the expected demand for AI-powered laptops.

While Intel has been busy making announcements, TheStreet Pro’s Stephen Guilfoyle noted that “it certainly seems that fewer folks care.”

Intel beat Wall Street’s first-quarter expectations for adjusted earnings but fell short of projections for revenue generation, he said.

Guilfoyle cited a 19.4% contraction in the data center and in AI as well as a 47.8% contraction in sales for Mobileye, an Israeli autonomous-driving company in which Intel has a controlling stake.

“When it comes to Intel, I am from Missouri,” Guilfoyle said. “They are going to have to show me.”

Story continues

“Advanced Micro Devices, Broadcom, Marvell Technology and Taiwan Semiconductor are all better at doing different things than Intel, which is really a jack of all trades at this point,” he added.

And speaking of AMD (AMD) , the semiconductor producer had a time at Computex Taipei 2024. The four-day technology trade show, operating under the theme of “Connecting AI,” wrapped up on Friday after attracting more than 85,000 ICT buyers and professionals.

During the event AMD introduced its Ryzen AI 300 Series processors with neural processing unit for next-generation AI personal computers and its next-gen AMD Ryzen 9000 Series processors for desktops.

Analyst says ‘AMD cannot be ignored’

AMD Ryzen AI 300 processors feature “the fastest APU performance, the world’s most powerful NPU with 50 TOPs, and the world’s first block floating point NPU doubling the performance of 16-bit applications without sacrificing accuracy,” Jack Huynh, AMD’s senior vice president and general manager for the computing and graphics group, said in a statement.

(Accelerated processing units handle both general processing and graphics tasks in one package. Neural processing units particularly manage AI tasks like speech recognition.)

AMD Chairwoman and Chief Executive Lisa Su showcased the growing momentum of the AMD Instinct accelerator family during the keynote.

More AI Stocks:

Raymond James reiterated its outperform rating for AMD, KLA Corp. (KLAC) and Lam Research (LRCX) on Friday, June 7.

Analyst Srini Pajjuri, who has a $180 price target on AMD, noted that management said the MI300 ramp at leading customers is “going well” and appears to have a line of sight to surpass its forecast of $4 billion in artificial-intelligence-related revenue this year.

Related: Analysts retool MongoDB stock price target after earnings

The company is working with multiple memory suppliers “and management believes that higher [high-bandwidth-memory] capacity gives the company an added [total-cost-of-ownership] advantage,” Pajjuri said.

“AMD is currently engaged with more than 100 customers for MI300 although a bulk of the $4B+ forecast for this year is driven by large customers.”

TD Cowen analyst Matthew Ramsay said that while Nvidia is the current market leader in AI processors, AMD cannot be ignored, according to Tipranks.

He said that Computex left him “more confident that Nvidia remains the leader in accelerated computing (as we recapped in naming Nvidia our Best Idea for 2024).”

Ramsay said he viewed AMD’s “strong forward roadmap and its progress on building its networking ecosystem as reinforcing our confidence it can serve as the alternative to Nvidia’s AI leadership (not to mention continuing to gain share in server with leadership general purpose central processing units).”

“Against a very large AI compute [total addressable market], we believe AMD is well on the way to cementing its position as the de facto alternative to Nvidia’s leadership position,” he said.

Related: Single Best Trade: Wall Street veteran picks Palantir stock