Table of Contents
Marvell said in its earnings call that demand tied to AI infrastructure deployments continues to accelerate spending on networking and optical connectivity, particularly as hyperscalers deploy larger AI clusters requiring more bandwidth and lower latency.
In sum – what to know:
Revenue growth – Marvell expects fiscal 2027 revenue to grow about 40% to nearly $11.5 billion, driven by data center demand tied to AI infrastructure deployments.
Networking surge – The company forecast more than 70% annual growth in its interconnect business as hyperscalers expand AI networking capacity.
AI expansion – Marvell said larger AI clusters, optical networking and custom AI silicon programs are driving stronger long-term demand.
Marvell Technology raised its AI infrastructure outlook after reporting stronger-than-expected demand across its data center networking, optical interconnect, and custom silicon businesses, with CEO Matt Murphy saying the company now expects quarterly revenue to reach $3 billion one quarter earlier than previously forecast.
The chipmaker reported fiscal first-quarter 2027 revenue of $2.418 billion, up 28% year-over-year and 9% sequentially. Marvell also guided second-quarter revenue to $2.7 billion, representing 35% annual growth.
“We are seeing strong demand and exceptional bookings across our entire data center portfolio,” Murphy said during the company’s earnings call.
Murphy noted that Marvell now expects fiscal 2027 revenue to grow about 40% year-over-year to nearly $11.5 billion, while fiscal 2028 revenue is projected to reach roughly $16.5 billion. The company had previously expected to hit $3 billion in quarterly revenue in Q4 of fiscal 2027, but now expects to reach that level in Q3.
The company said demand tied to AI infrastructure deployments continues to accelerate spending on networking and optical connectivity, particularly as hyperscalers deploy larger AI clusters requiring more bandwidth and lower latency.
Marvell expects its interconnect business to grow more than 70% this fiscal year, above prior projections of 50% growth. Murphy said networking is becoming increasingly critical as AI architectures evolve toward reasoning models, mixture-of-experts systems, and agentic AI applications.
“In agentic AI, a single user request may require agents to query AI models many times rather than just once,” Murphy said. “This substantially increases the volume of data traffic that must be transmitted and switched with very low latency.”
The executive also highlighted growing demand for scale-out and scale-across networking architectures linking multiple AI data centers. Murphy said larger AI clusters increasingly span multiple data centers because of power and space constraints, increasing demand for optical interconnect and data center interconnect (DCI) technologies.
Marvell said it expects its DCI business to reach a $1 billion annualized revenue run rate during fiscal 2028, roughly double fiscal 2026 levels.
Beyond networking, Marvell expects its custom silicon business to accelerate significantly next year. The company said custom revenue remains on track to grow more than 20% in fiscal 2027 and is expected to more than double in fiscal 2028, supported by existing XPU programs and new AI infrastructure deployments.
Murphy also said the company is planning for cloud capex growth to moderate into the “30% range” in fiscal 2028 while continuing to expect strong data center revenue growth.
“Our customers continue to signal robust demand, not only for this year, but for the next several years,” Murphy said.
The company added that it continues increasing R&D and capacity investments to support future AI infrastructure growth, particularly across optical networking, switching, and custom AI silicon platforms.
