Nvidia Commits $4 Billion to Optical Technology as AI Data Centers Hit Bandwidth Limits
Santa Clara, Monday, 2 March 2026.
The chipmaker’s massive investment in Lumentum and Coherent represents a strategic pivot toward photonics technology that uses light instead of electrical signals for data transmission. This move addresses the growing bandwidth crisis in AI infrastructure as advanced applications like agentic AI demand unprecedented data transfer speeds. The deals include multibillion-dollar purchase commitments and manufacturing capacity rights, positioning Nvidia ahead of competitors like AMD in the race for next-generation optical computing solutions.
Strategic Investment Details and Market Response
On Monday, March 2, 2026, Nvidia announced it will invest $2 billion each in Lumentum Holdings Inc. and Coherent Corp., both U.S.-based photonics companies [1][2][3]. The market responded positively to the announcement, with Lumentum shares jumping 8% and Coherent stock surging 13% following the news [2]. These multiyear strategic agreements include multibillion-dollar purchase commitments from Nvidia and future capacity access rights for advanced laser components [1][2]. The deals represent a significant commitment to securing optical technology supply chains as AI infrastructure demands escalate.
Understanding Photonics Technology in AI Infrastructure
Photonics technology harnesses light particles (photons) to create high-performance optical applications that can transmit data at significantly higher speeds than traditional copper cables [GPT]. Both Lumentum, headquartered in San Jose, California, and Coherent develop optical and photonic technologies including optical transceivers, circuit switches, and lasers used to move data at high speeds over long distances [1][3]. These systems offer substantial advantages over electrical signals, providing higher bandwidth, lower latency, and reduced power consumption [1]. The technology becomes critical as AI applications require massive data movement between graphics processing units (GPUs) in data centers.
Addressing AI’s Growing Bandwidth Crisis
The investment comes as agentic AI applications like Anthropic’s Claude Cowork and Microsoft’s Copilot Tasks drive up bandwidth requirements in AI data centers to support executing multiple tasks quickly [1]. Jensen Huang, Nvidia’s founder and CEO, emphasized the strategic importance, stating: ‘AI has reinvented computing and is driving the largest computing infrastructure buildout in history. Together with Lumentum, NVIDIA is advancing the world’s most sophisticated silicon photonics to build the next generation of gigawatt-scale AI factories’ [3]. This infrastructure challenge has become more pressing as Nvidia already leveraged its 2020 acquisition of network hardware company Mellanox to enhance NVLink technology and increase data flow between GPUs [1].
Competitive Landscape and Industry Investment
Nvidia’s $4 billion commitment positions the company ahead in the optical computing race, but competitors are also investing heavily in photonics [1]. AMD acquired silicon photonics startup Enosemi in 2025, which the company said would ‘accelerate’ AMD’s optics innovation for AI systems [1]. Additionally, the Defense Advanced Research Projects Agency (DARPA) issued a call for research proposals in February 2026 focused on improving photonic computing for AI applications [1]. Michael Hurlston, CEO of Lumentum, highlighted the collaboration’s significance: ‘This multiyear strategic agreement reflects our shared commitment to advancing the optics technologies that will power the next generation of AI infrastructure’ [3]. The investment includes support for expanding research and development capabilities and U.S.-based manufacturing, with Lumentum planning to invest in a new fabrication facility to increase capacity [3].