Enlightra raises USD 15 million to build energy-efficient laser links for AI infrastructure

Enlightra has secured USD 15 million to develop chip-scale, multiwavelength laser technology designed to improve the speed and energy efficiency of data transmission in AI clusters and data centers.

Lausanne-based deeptech start-up Enlightra has raised USD 15 million in funding to accelerate the development of chip-scale, multiwavelength laser technology designed to address one of the most pressing challenges in artificial intelligence infrastructure: fast, energy-efficient data transmission.

As AI models grow in size and complexity, data centers and AI clusters are facing mounting constraints linked to power consumption, heat dissipation and bandwidth limitations. Most high-performance computing systems still rely heavily on copper wiring to connect GPUs and accelerators, a technology increasingly seen as a bottleneck for both performance and sustainability. Enlightra aims to replace these electrical links with compact optical connections that deliver significantly higher bandwidth while consuming far less energy.

Founded in 2022, Enlightra develops multi-color, or “comb-laser,” technology that generates multiple optical wavelengths from a single chip-scale source. Each wavelength functions as an independent data channel, enabling dozens of high-bandwidth links to be transmitted simultaneously over optical fiber. According to the company, this approach reduces power consumption, cost and physical footprint compared with systems based on multiple discrete lasers.

The USD 15 million funding round has enabled Enlightra to design, build and demonstrate its laser technology, with initial 8- and 16-channel devices already meeting customer specifications for AI chip interconnects. The company reports error-free data transmission at target speeds and power levels, with pilot production planned for 2027. Investors include Y Combinator, Runa Capital, Pegasus Tech Ventures, Protocol Labs, Halo Labs, Asymmetry Ventures and TRAC VC.

“AI infrastructure is hitting the physical and energy limits of copper,” said Maxim Karpov, co-founder and co-CEO of Enlightra. “By using light as the backbone of GPU communication, we can decouple performance growth from energy consumption.”

Built using industry-standard silicon photonics fabrication processes, Enlightra’s lasers are designed for large-scale manufacturing, opening the door to deployment across global data centers. While AI clusters are the initial focus, the company sees broader applications in data center networking, subsea communications, chip-to-memory interconnects, and longer-term use cases such as quantum and space-based communications.