Nvidia has dominated the AI panorama since ChatGPT’s launch broke the web, kicking off a tidal wave of analysis into AI apps, together with chatbots and AI brokers.
Its graphic processing items, or GPUs, are ideally fitted to dealing with the heavy workloads related to coaching and working AI apps. Their means to course of knowledge quicker than conventional CPUs present in knowledge facilities has unleashed an enormous knowledge middle improve cycle, leading to a whole lot of billions of {dollars} in gross sales for Nvidia.
Nvidia’s moat, nevertheless, could also be shrinking as different chipmakers refine their very own semiconductors to go well with AI necessities higher. Among the many semiconductor shares furthest alongside in that journey is Marvell Know-how (MRVL). The corporate, based in 1995, made a significant pivot towards knowledge middle infrastructure in 2016 underneath the management of CEO Matt Murphy, a transfer that positioned it completely to seize the expansion in AI spending.
Marvell’s experience in creating application-specific built-in circuits, or ASICs, that effectively carry out particular, routine workloads offered it with the instruments wanted to companion with corporations like Amazon to construct {custom} AI chips, referred to as XPUs, as hyperscalers looked for Nvidia alternate options to diversify their provide chains and scale back prices.
These partnerships have offered a pleasant jolt to Marvell’s gross sales and revenue progress, and following shows at Amazon AWS’s latest re:Invent convention, Marvell is on the cusp of a significant step up in demand as demand for XPUs and interconnect merchandise used to tie networks collectively climbs.
Marvell carves profitable area of interest towards Nvidia
Nvidia’s chips stay the go-to alternative for hyperscalers and knowledge facilities as a result of they’re quicker, general-purpose options extremely optimized for AI duties, due to Nvidia’s CUDA software program. The corporate controls over 80% of the AI chip market, and gamers like Marvell are unlikely to displace its dominance.
Marvell Know-how is experiencing progress as Amazon deploys extra {custom} silicon and knowledge facilities want for interconnects improve.
Shutterstock&interval;
That mentioned, Marvell is prone to carve away billions of {dollars} in income that might in any other case have gone to Nvidia as Amazon continues to take a position closely in creating its Trainium chip lineup.
“Effective Tuesday, Amazon launched its Trainium3 chip, which is part of the revenue ramp called out by Marvell back in August, and that program, along with others, should help drive Marvell’s custom AI silicon business higher over the coming quarters,” mentioned long-time portfolio supervisor Chris Versace in a put up on TheStreet Professional.
Amazon Trainium chips supply benefits over GPUs:
- Cheaper and extra environment friendly: Trainium chips are custom-built to coach machine studying fashions, resembling AI’s massive language fashions. Amazon claims that its use can scale back coaching prices by 50% in comparison with GPU-based methods. The chips are particularly designed for Amazon’s total knowledge middle infrastructure, making certain most optimization in comparison with off-the-shelf GPU options.
- Diversification: Concentrating knowledge facilities round GPUs exposes Amazon to produce chain dangers and limits negotiating energy, rising Amazon’s prices at the same time as they’re already surging.
- Creates extra income streams/stickiness: Trainium chips are proprietary to AWS, suggesting that enterprises’ use of them deepens relationships and will increase buyer switching prices, all of the whereas offering AWS with a brand new income stream tied to utilization.
Rising deployment of Trainium chips inside Amazon’s AWS is already supporting income progress at Marvell’s knowledge middle enterprise. Within the third quarter, knowledge middle gross sales, primarily pushed by AI merchandise together with XPUs and interconnects, totaled $1.52 billion, representing a 38% improve from the identical interval final yr and accounting for the lion’s share of Marvell’s whole income of $2.07 billion.
Customized XPU gross sales have been $418 million within the quarter, up 83% yr over yr.
Amazon’s latest Trainium3 chip is much more highly effective and environment friendly. Its Trainium3 UltraServers are as much as 4 occasions extra power environment friendly and possess 4 occasions the reminiscence of its Trainium2 UltraServers.
“We are guiding for robust growth in the fourth quarter and are on track for a strong finish to the fiscal year, with full-year revenue growth forecasted to exceed 40%. Looking ahead, we see demand for our products continuing to accelerate, and as a result, our data center revenue growth forecast for next year is now higher than prior expectations,” said Matt Murphy, Marvell’s Chairman and CEO.
What’s next for Marvell Technology?
CEO Murphy provided solid guidance for the current quarter, stating that revenue is expected to be around $2.2 billion, up from $1.8 billion in the same period last year.
Next year could be even better, given the significant investment Amazon is making in building additional data center capacity for its customers, including Anthropic, whose AI chatbot, Claude, is among the most popular.
Amazon has invested about $8 billion in Anthropic to fuel its growth, and unsurprisingly, Anthropic has committed to using Trainium chips to train its models.
“Trainium2, it is actually doing properly. It is absolutely subscribed on Trainium2. We’ve — it is a multibillion-dollar enterprise at this level. It grew 150% quarter-over-quarter in income,” said Jassy on the earnings call. “We’ve a number of demand for Trainium.”
Amazon’s capital expenditures surged to $125 billion this year due to its significant investments in AI, including Trainium. In Q3, it spent $34 billion, about $10 billion more than it spent in Q1, 2025. Last year, its capex was $83 billion.
That spending isn’t expected to slow, given Amazon CFO Brian Olsavsky said on its third quarter earnings call, “We count on that quantity will improve in 2026.”
The spending will help AWS deliver on its latest plan to build “AI Factories” that can be deployed onsite for non-cloud enterprise and government use. Those factories will include Trainium chips and Nvidia GPUs. They’ll also need plenty of interconnect products, like switches, active electrical cables, transceivers, and amplifiers, further supporting Marvell’s sales growth, given that interconnect represents about half of Marvell’s data center sales.
Further out, Marvell Technology expects to begin producing XPUs for a second hyperscaler client, with meaningful revenue expected to be generated in the next couple of years.
In a research note shared with TheStreet, Morgan Stanley analysts said that Marvell’s guidance includes custom silicon growth of 20% in 2026 and 100% in 2027.
“We count on {custom} progress subsequent fiscal yr to be increased within the second half and don’t count on any air pockets in {custom} income,” said Murphy. “We count on accelerated progress over the subsequent a number of years. fueled by our rising portfolio of design wins.”
Morgan Stanley upped its Marvell stock price target to $112 from $86. Meanwhile, Chris Versace’s price target increased to $140 from $125.
Associated: Goldman Sachs points Micron prediction forward of earnings
