Xscape Photonics secures $37M to rethink AI data center power
Xscape Photonics, a photonics startup focused on next‑generation data center infrastructure, has raised $37 million in fresh funding to accelerate the rollout of its laser‑based interconnect platform designed specifically for AI data centers.
The investment will help the company industrialise its flagship product, the FalconX platform, which uses multi‑wavelength laser technology to move data more efficiently between servers and accelerators. As generative AI workloads surge, operators are struggling with escalating energy consumption, network congestion and the physical limits of traditional copper and standard optical links.
FalconX: 8‑color laser engine for high‑density AI clusters
The core of Xscape Photonics‘ value proposition is its 8‑color FalconX engine, which employs multiple laser wavelengths on a single link to dramatically increase bandwidth while cutting power use and heat. By integrating advanced silicon photonics with custom packaging, FalconX is engineered to sit closer to GPUs and AI accelerators, reducing latency and improving overall system throughput.
According to the company, this architecture enables higher‑density AI clusters without proportionally increasing cooling and power requirements, a critical constraint for hyperscale and enterprise data centers racing to deploy larger AI models. The platform is targeted at cloud providers, large internet companies and high‑performance computing sites that are upgrading to next‑generation AI infrastructure.
Rising demand for energy‑efficient AI networking
As spending on AI infrastructure rapidly climbs, industry analysts warn that networking and interconnects are becoming a bottleneck. Traditional approaches cannot easily keep pace with the explosive growth in model parameters, training data and GPU counts. Solutions based on photonics and optical interconnects are increasingly seen as essential to sustain performance while keeping energy use in check.
With its new funding, Xscape Photonics plans to expand engineering, scale manufacturing partnerships and run large‑scale trials with leading cloud and AI customers. If successful, its FalconX technology could become a key building block in the next wave of AI data center design, where bandwidth, latency and power efficiency determine competitive advantage.