Marvell's Latest Tech Leap with TSMC - A 2nm Chip Platform Set to Transform AI Infrastructure

Marvell Technology Inc MRVL has unveiled a significant development in semiconductor technology through an expanded partnership with Taiwan Semiconductor Manufacturing Co TSM, launching the first-ever 2nm silicon intellectual property (IP) technology platform. 

This pioneering platform is tailored for acceleration infrastructure applications, including Artificial Intelligence (AI) clusters and cloud data centers. 

Marvell’s announcement underlines the creation of an IP portfolio that serves a broad spectrum of infrastructure needs, the Digitimes reports. 

Also Read: Nvidia Expands Dominance with Custom AI Chip Design, Challenging Broadcom and Marvell’s Territory

It incorporates technologies such as high-speed SerDes capable of exceeding 200 Gbps, processor subsystems, encryption engines, SoC fabrics, and various high-bandwidth physical layer interfaces. 

The collaboration promises to significantly boost chip performance, power efficiency, and area optimization (PPA), which are vital for handling advanced acceleration workloads. 

Although Marvell has not yet announced the silicon certification for its 2nm IP platform, the collaboration with TSMC on N2-compatible IP hints at the potential readiness of Marvell’s leading IP solutions. 

Last week, Marvell reported a fourth-quarter revenue growth of 1% year-over-year to $1.427 billion, surpassing the consensus estimate of $1.42 billion. The chip designer reported an EPS of $0.46, which aligns with analyst estimates. 

Marvell also boosted its share buyback by $3 billion to $3.3 billion.

Investors can gain exposure to Marvell via IShares U.S. Digital Infrastructure And Real Estate ETF IDGT and Spear Alpha ETF SPRX.

Price Action: MRVL shares traded higher by 2.42% at $67.69 on the last check Friday.

Also Read: Nvidia Supplier Taiwan Semi’s Global Expansion and AI Revenue Growth in Focus Despite Overbought Signals

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo via Shutterstock