AMD Says MI300 Is Its ‘Fastest-Ramping Product,’ Teases New AI Chips ‘Later This Year’

In AMD's first-quarter earnings call, CEO Lisa Su teases that the company will share details in the 'coming months' about new Instinct accelerator chips for data centers arriving 'later this year into 2025' amid heightened competition with Nvidia.

AMD’s business is in much better shape than it was a year ago, thanks to its PC and data center revenues growing more than 80 percent each over that period. Helping with that growth was its Instinct MI300 AI chips, which CEO Lisa Su called the “fastest-ramping product” in its history.

The Santa Clara, Calif.-based company reported on Tuesday that this year-over-year growth within its two largest business segments happened in the first quarter, which ended March 30.

Total revenue for the quarter was $5.5 billion. While this represented an 11.3-percent sequential decline, it marked a 2.2 percent increase from the same period last year when AMD’s revenue took a massive hit due to a nosedive in PC chip sales and a slowdown in data center chip sales.

On AMD’s earnings call, Su said the company’s solid performance in the PC and data center segments sets it up with enough forward momentum to return to “strong annual revenue growth and expanded gross margin” this year. The last time AMD’s annual revenue grew was in 2021.

“Our priorities for 2024 are very clear: accelerate our data center growth by ramping Instinct GPU production and gaining share with our Epyc processors, launch our next generation Zen 5 PC and server processors that extend our leadership performance and expand our adaptive computing portfolio with differentiated solutions,” she said.

AMD expects revenue in the second quarter to reach roughly $5.7 billion, plus or minus $300 million. At the midpoint, this would amount to an approximately 3.6 percent sequential increase and about 5.6 percent higher than the same period last year.

The company’s first-quarter revenue was slightly higher than Wall Street’s expectations while its second-quarter revenue forecast was at the midpoint of analysts’ predictions.

AMD’s shareholders were apparently not impressed enough, sending the company’s stock price down by more than 7.7 percent in after-hours trading Tuesday.

Su: Data Center GPU Sales To Hit $4B, Teases Future Chips

With AMD looking to challenge Nvidia’s dominance in the AI computing space, Su said expanded customer engagements with the company’s Instinct MI300 accelerator chips prompted it to upgrade its 2024 data center GPU revenue forecast to $4 billion.

This represents a $500 million bump from the data center GPU guidance the company offered in January. In the month prior, AMD marked the official launch of the MI300 chips as a challenge to Nvidia’s powerful and popular H100 GPU.

Su said more than 100 enterprise and AI customers are “actively developing or deploying” the MI300X, the GPU-only version of the processor. (The other version is the MI300A, which combines GPU cores and CPU cores onto a single package.)

What’s helping with demand are the multiple MI300X systems entering volume production this quarter with OEMs such as Dell Technologies, Hewlett Packard Enterprise, Lenovo, Supermicro and others, according to Su. Cloud service providers Microsoft and Oracle have also expanded their MI300X production environments, as did Facebook parent company Meta.

“What we see now is just greater visibility to both current customers as well as new customers committing to MI300,” she said.

This increased interest helped the MI300 become AMD’s “fastest-ramping product” in its history, generating more than $1 billion in “total sales in less than two quarters.”

Su suggested that AMD could sell even more MI300 products if it wasn’t for supply constraints that were holding the company back, which she considered an industry-wide issue. The company expects MI300 supply to “improve every quarter this year.”

“We are tight on supply, so there’s no questions in the near term that if we had more supply, we have demand for that product, and we're going to continue to work on those elements as we go through the year,” Su said, adding that she’s “very pleased with how the ramp is going.”

While Su said she is satisfied with AMD’s progress with the MI300, the company has to contend with the fact that Nvidia recently started shipping the successor to its H100, the memory-rich H200. The rival also plans to launch more powerful follow-up chips later this year that use the company’s next-generation Blackwell architecture.

Asked about Nvidia’s accelerated release cadence for AI chips, Su said AMD’s strategy is not about one product but a “multi-year, multi-generational road map” and teased that it will share details in the “coming months” about new accelerator chips arriving “later this year into 2025.”

“We’re very confident in our ability to continue to be very competitive. Frankly, I think we’re going to get more competitive,” she said.

Su Expects Things To Get Better For PC, Data Center Businesses

Su said AMD expects its PC business to return to annual growth this year, thanks to the “start of an enterprise refresh cycle and AI PC adoption,” which the company plans to address with its expanding portfolio of Ryzen processors for laptops and desktops.

For the first quarter, client revenue was $1.4 billion, a 6.4 percent sequential but an 85 percent increase from the same period last year when PC chip sales plummeted.

“We see AI as the biggest inflection point in PCs since the internet, with the ability to deliver unprecedented productivity and useability gains,” she said.

The chief executive sees “clear opportunities to gain additional commercial PC share based on the performance and efficiency advantages of our Ryzen Pro portfolio and an expanded set of AMD-powered commercial PCs from our OEM partners.”

AMD expects its Ryzen portfolio to gain even more traction later this year when it launches a next-generation series of mobile processors with the code name “Strix.”

“Customer interest in Strix is very high based on the significant performance and energy efficiency uplifts we are delivering,” Su said. “Design win momentum for premium notebooks is outpacing prior generations as Strix enables next-generation AI experiences in laptops that are thinner, lighter and faster than ever before.”

The company’s data center business saw a similar recovery in the first quarter from last year with $2.3 billion in revenue. This represented a 2.4 percent sequential increase and an 80 percent increase from the same period last year.

While AMD’s MI300 contributed a significant amount of growth to the data center business, Su said the company also saw a “double-digit percentage increase in server CPU sales.”

“In the enterprise, we have seen signs of improving demand as CIOs need to add more general purpose and AI compute capacity while maintaining the physical footprint and power needs of their current infrastructure,” she said.

The chief executive teased that AMD’s next-generation Epyc server CPUs, code-named Turin, will use the company’s Zen 5 architecture and provide “significant performance and efficiency increases” for cloud computing workloads when they launch “later this year.”

The two areas where AMD struggled were its embedded and gaming businesses.

Embedded revenue, which includes products from AMD’s Xilinx acquisition, was $846 million, a 20 percent sequential decline and a 45.8 percent year-over-year decline.

“Longer term, we see AI at the edge as a large growth opportunity that will drive increased demand for compute across a wide range of devices,” Su said.

Gaming revenue, on the other hand, declined sequentially by 32.6 percent and 47.5 percent year-over-year to $922 million, due to lower demand for PC GPUs. AMD’s CFO, Jean Hu, said the company doesn’t expect the situation to improve this year.

“Based on the visibility we have, the first half […] we guided down sequentially more than 30 percent. We actually think the second half will be lower than the first half,” she said.