Ayar Labs, which is developing a way to link AI chips directly with optical connections, raised a $150M extension to its $500M Series E announced in March 2026
Ayar Labs, a Silicon Valley startup developing a way to link AI chips directly together with optical connections, on Thursday said it has raised an additional $150 million.
Context & Ripple Effects
Ayar Labs has progressed from a 2020 Series B for high-performance optical interconnects to a $130 million Series C backed by Nvidia, Intel and others in 2022. Its March 2026 Series E framed the technology as a replacement for copper wiring in semiconductors; the extension takes that financing round to $650 million.
The raise lands amid a broader contest to relieve AI-system data-transfer constraints: Celestial AI raised $175 million for optical processor-to-memory links in 2024, while Eliyan raised $145 million for chiplet data-transfer technology in July 2026. The scale of Ayar’s round makes funding capacity part of the competition over interconnect architecture.
First-order effects
- Ayar Labs receives an additional $150 million to advance optical links that directly connect AI chips, extending the $500 million Series E announced in March 2026.
- The extension gives Ayar a larger financial base than was established by its March round as it pursues replacing copper semiconductor wiring with fiber optics.
Second-order effects
- Celestial AI and Eliyan face a better-funded rival in efforts to address data-transfer bottlenecks between AI-system components, increasing pressure to differentiate their optical-link and chiplet approaches.
- Investors and AI-hardware partners gain a clearer signal that optical interconnects are attracting late-stage infrastructure capital rather than remaining confined to earlier-stage R&D funding.
Third-order effects
- If AI-chip clusters continue to be constrained by connections between components, interconnect design may become a more consequential layer of AI hardware, alongside processors and memory.
- The pattern points toward AI infrastructure funding reaching specialized semiconductor subsystems whose value depends on deployment at larger system scale.
The trend: AI infrastructure capital is moving deeper into optical and chip-to-chip interconnects as AI systems make data movement a central hardware constraint.