Nvidia The company has spent at least $6.5 billion in companies developing photonics technology in the past three months as it rushes to invest to solve one of the major bottlenecks in AI deployment.
Photonics, which uses light to transmit data, is an emerging technology that is considered a more efficient alternative to current processes that use electricity to transmit data. Electrical data transfer consumes more energy. This factor is increasingly seen as an impediment to the widespread deployment of AI.
Since early March, Nvidia has announced $2 billion in investments. lumen, reasonable and marvelboth of which are developing photonics technology. The semiconductor giant also announced an investment of $500 million. corning Participated in optical startup Ayer Labs’ $500 million Series E funding round to develop advanced optical connectivity solutions.
“Photonics is a way for Nvidia to scale its AI infrastructure without incurring the energy costs associated with using power and copper,” Forrester senior analyst Alvin Nguyen told CNBC.
“By investing in photonics companies, Nvidia ensures that advances in photonics continue and prevent us from hitting the scalability and performance walls that would occur if we stayed with electricity and copper.”
Resolving bottlenecks
Photonics can be used in AI infrastructure by using light to move data between graphics processing units (GPUs), memory, networking chips, servers, and data centers, rather than relying solely on electrical signals flowing along copper wires.
Brian Colello, senior equity analyst at Morningstar, told CNBC that while copper is currently the dominant connectivity standard due to its low cost and high reliability, photonics will become more prominent in AI infrastructure over time.
“NVIDIA’s next-generation AI rack-scale solution roadmap requires an increase in the amount of optical connectivity to handle exponentially increasing bandwidth due to new models and increased usage,” he said.
The semiconductor giant already makes some photonics technologies available as part of its networking solutions offering, and announced tools that will allow AI factories to connect millions of GPUs across sites while significantly reducing energy consumption and operating costs.
“If we look upstream, we come to the conclusion that we are starting to scale silicon photonics technology,” Nvidia CEO Jensen Huang said at GTC in March, pointing to Nvidia’s Ethernet networking platform used to connect AI factories and GPU clusters. He also said the company is starting to add photonics to the interconnect technology between GPUs.
“This means that the amount of silicon photonics technology capability we need is significantly greater than the amount the world has today,” he added. “So we are working with the supply chain to be able to help them ramp up production capacity in advance of that.”
Lumentum stock is up 134% since the beginning of the year, while Coherent is up 96%. Marvell’s stock price rose 122% in 2026, while Corning’s stock rose 111%.
Stock prices of photonics companies have soared over the past year.
Nvidia is one of many AI players that has recently moved to pour money into photonics technology.
chip manufacturer AMD It joined Nvidia in the Ayer Labs round, acquired startup Enosemi in 2025, and also made equity investments in Teramount and Celestial AI. alphabet and microsoft Venture Arms backed nEye with an $80 million Series C in April.
However, deploying photonics technology at scale across the AI infrastructure stack comes with its own challenges.
“The technology is sound, but the scale of production is the bigger challenge,” Nick Payens, head of AI at Futurum Group, told CNBC.
“Manufacturing yields for complex co-packaged optical assemblies remain a challenge, as the precise alignment of optical and silicon components is unforgiving and if something goes wrong during the packaging process, it is usually not possible to redo the assembly,” he said.
“The transition is underway, but it’s still in its early stages,” Patience added. “We expect large-scale adoption to occur in 2028 and beyond.”
