Marvell Technology (MRVL) Q1 FY27 earnings call: guidance and Q&A

Confirming the $10 billion+ custom XPU revenue target for FY29 and asking when investors can expect more details or credit for the new large XPU customer program, given its potential exclusivity and large CapEx profile.

Matt Murphy confirmed the $10 billion+ custom silicon revenue target for FY29, noting it was an update from a prior $8 billion target based on a larger TAM. He stated the company is on track, driven by existing programs, new ramps, and XPU attach programs. He expects confidence in the new program's magnitude to grow as the year progresses, emphasizing it's a key part of the plan but not the sole driver, representing about one-third of custom growth next year.

“Yes. So yes, you heard it right. And just for context on the $10 billion plus for fiscal '29, the context for that is, back in April 2024 we set long-term targets out through calendar '28, which would be our fiscal '29.”

asked by Vivek Arya · answered by Matthew Murphy

Asked about Marvell's leverage of its SRAM-based IP differentiation in XPU offload engines, given the aggressive evolution of inference workloads and the company's history with SRAM design capabilities.

Matt Murphy confirmed Marvell's strong SRAM design capability and IP, stemming from a long legacy including the Avera acquisition. He stated this is a key part of their IP portfolio and a reason for winning XPU attach designs, noting it's part of a broader strategy including advanced packaging, high-speed I/O, and manufacturing expertise. He acknowledged the trend towards SRAM-based offload architectures and will provide updates.

“That technology has evolved from networking products into AI products and you can see that trend in the market as inferencing is happening. So that continues to be a key part of our IP portfolio sure as we go win designs. And it's one of the reasons why we are winning our designs in XPU attach.”

asked by Harlan Sur · answered by Matthew Murphy

Inquired about the breadth of Marvell's custom customer base beyond existing XPU and XPU attach programs, specifically asking if moving into the compute TAM (the largest custom compute wallet) is included in the forecast or would be incremental.

Matt Murphy stated that Marvell has custom engagements across all U.S. hyperscalers, including XPU and XPU attach. He clarified that the current growth outlook (20%+ this year, doubling next year, and long-term targets) is based on designs already won, with newer wins acting as an 'insurance policy' for later contributions. He highlighted the competitiveness of Marvell's technology platform, especially its high-speed I/O and SerDes performance, which is driving new opportunities and strengthening their roadmap.

“So everything we've laid out, which is the greater than 20% growth this year, more than doubling next year and then still having line of sight to our long-term targets. That's all based on the designs we've already won and locked and even going back really the last summer.”

asked by Timothy Arcuri · answered by Matthew Murphy

Asked about the reality of the CXL opportunity alongside accelerators and how to quantify the content per accelerator.

Matt Murphy affirmed that CXL is a very real and expanding opportunity, noting Marvell engaged in this business years ago for traditional server dynamics, which has now vectored into AI. He mentioned that the plan to get the custom XPU attach line item to over $1 billion in revenue in the next couple of years continues to expand due to customer base growth and concerns around memory cycles driving CXL adoption. He indicated that memory architectures are critical for next-generation infrastructure, playing in Marvell's favor.

“I think we had a plan in this area, which we've talked about getting that sort of custom line item of XPO attach alone over $1 billion in revenue in the next couple of years. That continues to expand both because we continue to expand the customer base, which is very compelling.”

asked by Blayne Curtis · answered by Willem Meintjes

Asked for more color on how Marvell is addressing capacity constraints and whether the increased guidance is due to securing more capacity or increased confidence in customer forecasts.

Chris Koopmans explained that Marvell manages capacity by building tight relationships with key suppliers, providing 5-year forecasts, and backing those forecasts with strategic prepayments. He stated that the AI infrastructure build-out has been constrained since its beginning, and Marvell's approach helps them secure necessary capacity. Matt Murphy added that the supply chain operations have done an outstanding job reacting to upward demand changes, putting them in great shape to deliver on their outlook.

“And ultimately, the way we've been able to manage this is by building very tight relationships with a small number of key suppliers, giving them a 5-year forecast of what we need and hitting what we need each time along the way and doing what we said we were going to do.”

asked by Chris Caso · answered by Christopher Koopmans

Questioned why the interconnect business growth, which is accelerating (70%+ in FY27), is projected to slow down in FY28, even with strong tailwinds like DCI, AEC, and 1.6T.

Matt Murphy acknowledged the strong acceleration in interconnect growth and stated that while the current FY28 outlook is where they are today, there's definite potential for upward bias. He cited traditional DSP business (1.6T step-up), DCI ramping, new initiatives like retimers and AECs, and scale-up optics (projected $300 million next year) as areas that could drive further growth. He emphasized that there's a lot of optionality and upward bias for the interconnect business.

“I think there's a lot of optionality is what I would say at the moment, Ross. But I think today, net-net, you look overall, 16.5 where we're comfortable overall, but I think there's upward bias for sure. The trends continue.”

asked by Ross Seymore · answered by Matthew Murphy

Asked to highlight 2-3 specific products or dynamics within the interconnect business, especially scale-up, that are contributing most to the new growth forecast, given the dynamic market and various technologies (copper, optical, NPO, CPO).

Matt Murphy highlighted Marvell's unique advantage of having the broadest range of connectivity solutions across all technologies. He pointed to the intense activity and success of combining the Celestial AI team with Marvell's optics team, enabling full end-to-end solutions from XPU to switch. He emphasized that this comprehensive approach, including Marvell's proven silicon photonics technology (15 billion hours of field data), 224 gig SerDes, and high-yield switch manufacturing, is resonating with customers for scale-up networks. He sees this end-to-end capability as a major growth driver with significant future potential.

“The good news is, as you said, you just rattled off a number of different technologies, and there's an equal number more you didn't mention. We're in all of them. And I think that's one of the unique advantages that we bring to the table is we have the absolute broadest range of connectivity and scale up and scale out solutions for interconnect in the industry.”

asked by Tore Svanberg · answered by Matthew Murphy

Asked for more insight into the switching side, particularly the scale-up opportunity, given Marvell's relatively small share in the large scale-out market and the greenfield nature of scale-up with multiple protocols (ESUN, NVLink, UALink).

Matt Murphy noted that reaching $1 billion in scale-out switch revenue next year is a huge milestone, leveraging assets from the Innovium acquisition. He sees scale-up switching as a bigger opportunity because the market share isn't established, and it requires a broad set of capabilities and track record. He emphasized that scale-up switching is not significantly factored into the current $16.5 billion FY28 revenue outlook, representing substantial upside. He believes Marvell's assets and strategy position them well to establish leadership in this emerging market.

“So that, in our mind, is a bigger opportunity from the standpoint that the market share isn't established yet -- and that's what I was referring to earlier in Tore's question, there's so many important architectural discussions going on, but it's not just the switch, and it really requires kind of a very broad set of capabilities and track record for customers to bet on you here.”

asked by Srinivas Pajjuri · answered by Matthew Murphy