Marvell stock pops on report it will help Google with custom AI chips

Shares of Marvell Technology gained nearly 6% on Monday. reports He said Google will use the chip design firm for two new chips to power artificial intelligence workloads.

Until now, Google has relied on Marvell rival Broadcom to design its in-house Tensor Processing Units, or TPUs. Broadcom shares fell nearly 2% on Monday following the report. Information.

The potential deal between Google and Marvell could include a memory processing unit as well as a TPU, The Information reported Sunday. Google and Marvell did not immediately respond to requests for comment.

Both Marvell and Broadcom help their customers translate chip designs into silicon, providing back-end support before processors are shipped to large manufacturing facilities by companies like Taiwan Semiconductor Manufacturing Company.

It’s been a role fueling the growth of both Marvell and Broadcom as more tech giants design in-house accelerators for AI.

Amid a push to produce enough silicon to power AI, it’s no surprise to see Google diversifying its chip deals beyond Broadcom. The Google-Broadcom partnership is alive and well, having been extended through 2031 in an expanded agreement announced earlier this month.

Last week, Meta also struck a big deal with Broadcom, committing to deploying its own 1-gigawatt custom MTIA chip using Broadcom technology.

Marvell shares gained more than 20% in March as the company posted strong fourth-quarter earnings and guidance amid rising demand for artificial intelligence. Shares have continued their rise in April, rising almost 50% so far.

Nvidia It also announced a $2 billion investment in Marvell in March. The deal makes it easier for Nvidia customers to access application-specific integrated circuits, or ASICs, built by hyperscalers like Google.

Google was the first hyperscaler to begin developing its own custom ASIC to accelerate AI workloads, releasing its first TPU in 2015. As Big Tech scrambles to find adequate compute and lower-cost alternatives to Nvidia’s AI chips, giants like Amazon, Meta, Microsoft, and OpenAI have all followed suit.

Google released its latest 7th-generation “Ironwood” TPU in November and may release its next chips at its annual AI conference, Google Cloud Next, later this week.

Originally trained for internal workloads, Google’s custom microchip has been available to cloud customers since 2018. Meta, Anthropic, and Apple all now use TPU, as Google increasingly encroaches on a market cornered by Nvidia’s graphics processing units.

Memory has been one of the many bottlenecks AI chip makers have faced in recent months due to supply shortages from memory makers such as Micron, SK Hynix and Samsung.

CNBC’s Kristina Partsinevelos contributed to this report.

Watch: Inside Google’s chip lab producing custom silicon to train Gemini and Apple AI models