Eliyan Series C raises $145M at $1B valuation for AI interconnects
Eliyan raised $145 million at a $1 billion valuation to build optical interconnects for AI data centers, with Seligman Ventures leading.
By Colin Brandt · Enterprise Reporter
· 3 min read
Eliyan Corp. said its Eliyan Series C brought in $145 million and lifted the Silicon Valley chip interconnect startup to a $1 billion valuation. Seligman Ventures led the round, with Cisco Ventures and Lumentum participating, as investors continue to fund infrastructure bets tied to the bandwidth and power limits of large AI clusters.
The company is building electro-optical interconnect technology meant to move data across compute, memory and networking systems in AI data centers. Eliyan says its products are aimed at a specific bottleneck: as operators assemble larger pools of AI accelerators for large language models and agentic applications, existing interconnect and power approaches can leave chips waiting on data rather than running at full utilization.
Eliyan did not disclose revenue, headcount, customer names or the ownership terms attached to the round. It also did not name the hyperscalers, AI accelerator makers or memory suppliers that Chief Executive Ramin Farjadrad said are evaluating or deploying its technology.
What does Eliyan make?
Eliyan makes chip connectivity technology, including NuLink PHYs and NuGear chiplets, for die-to-die, chip-to-chip and rack-to-rack links. The company says the approach can increase bandwidth and improve power efficiency across distributed AI systems, including co-packaged computing fabrics.
That puts Eliyan in a less visible but increasingly contested part of the AI infrastructure stack. GPUs and custom accelerators get most of the attention, but cluster performance increasingly depends on how quickly data can move among processors, memory and networking components. If those links become constrained, training and inference workloads can stall even when expensive compute is installed.
Farjadrad said in a statement that AI is forcing data center architecture to change as compute, memory, packaging and networking become more tightly connected. He said Eliyan was created to address those constraints at the system architecture level, and that the new capital will support product development, customer adoption and ecosystem expansion.
The company said it plans to spend the funding on research and development, faster interconnect technology, manufacturing capacity and broader chiplet compatibility through partners. Those plans are directionally clear, but Eliyan did not provide production volumes, timelines for new products or named design wins.
Why optical interconnects matter for AI clusters
Optical interconnects use light-based signaling for high-speed data movement, a technique that can reduce some of the bandwidth and power penalties associated with moving data across dense computing systems. In AI data centers, that matters because the cost of a cluster is tied not only to the accelerators but also to whether the surrounding fabric can keep them supplied with data.
Seligman Ventures managing partner Umesh Padval pointed to Eliyan’s technical depth and portfolio of more than 100 patents. He said optical interconnects are becoming a performance-defining technology for AI data centers, even if the category is less prominent than processors or model software.
The round gives Eliyan unicorn status in a hardware category where capital requirements are usually higher and commercialization cycles are longer than in software. The disclosed valuation suggests investors are assigning value to the interconnect layer as AI infrastructure spending shifts from accelerator procurement alone toward full-stack cluster performance.
This story draws on original reporting from SiliconANGLE.
