Ayar Labs Adds $150M—but Its Optical Interconnect Deadline Is 2027

San Jose-based Ayar Labs secured an additional $150 million on September 10, 2026, bringing its primary capital raised during the year to $650 million, according to the company’s September 10 funding release. The same release assigns the capital to product development, validation, expansion of the manufacturing ecosystem and a new chip-design center in Bengaluru, India.

The financing is not evidence that Ayar’s optical interconnects are ready for mass production. A Reuters account of the transaction identifies the new money as an extension of a $500 million Series E round, distinguishes it from a separate $225 million secondary share purchase valuing Ayar above $5 billion, and places volume-production qualification at the end of 2027 before expected customer ramps in 2028–2029.

Primary financing and secondary liquidity serve different purposes

The expanded Series E is primary capital: money entering Ayar that can be used for engineering, validation and preparation of its supply chain. Combining the original round with the extension produces the company’s stated primary-funding total for the year.

The secondary transaction instead involved Antero Peak Group at Artisan Partners buying shares from early employees and investors, with Sequoia Global Equities, ARK Invest and Greycroft participating. Its proceeds went to the selling shareholders rather than into Ayar’s operating budget. Adding that purchase to the primary total would therefore overstate the resources available for product development and manufacturing.

The valuation attached to the secondary sale measures what those buyers were prepared to pay for existing equity. It does not establish that Ayar has completed component qualification, achieved repeatable manufacturing yields or secured high-volume orders.

The new capital targets manufacturing readiness

A working optical link is only one stage between laboratory development and a component that customers can place in production systems. Silicon design and optical performance must be supported by repeatable fabrication, packaging, testing and quality-control processes. Customers then need qualified components that can pass through their own integration and product-approval programs.

The Bengaluru center expands Ayar’s engineering presence alongside teams in the United States and Hsinchu, Taiwan. Its remit covers end-to-end silicon product development, while the wider funding plan includes validation and deeper integration with the semiconductor manufacturing ecosystem.

Those activities explain why additional capital and production readiness are separate claims. Funding can pay for engineering capacity, test work and supplier coordination, but it does not demonstrate that qualification has been completed. The visible commercial milestone is a component that can be manufactured consistently and accepted into a customer’s product process.

TeraPHY must move from technical promise to deployed hardware

Ayar develops co-packaged optics for AI scale-up systems, where accelerators and networking silicon require high-speed connections. SiliconANGLE’s product coverage identifies TeraPHY as Ayar’s optical engine and describes its placement close to an ASIC or GPU, allowing electrical signals to be converted to light near the compute package.

The architectural aim is to avoid carrying the fastest signals over longer copper connections, where reach, power demand and heat become harder to manage as clusters expand. These are the technology’s intended advantages, not proof of broad commercial deployment. The available disclosures do not identify a high-volume customer shipment or provide production-yield data.

Ayar’s collaboration with Taiwanese cloud-infrastructure provider Wiwynn offers a path toward system integration. It combines Ayar’s optical engines and remote light-source technology with rack-level server architecture, demonstrating ecosystem work beyond an isolated component. It does not, by itself, establish the timing or scale of a commercial rollout.

Qualification must precede the customer ramps

The production sequence now has a clear order: expanded financing and engineering capacity, component qualification, customer integration and then manufacturing ramps. The expected ramp window belongs to customers’ products rather than to a standalone Ayar launch, which makes it a planning horizon rather than a confirmed release schedule for named hardware.

The qualification target is the operative deadline. A delay would leave customers less time to integrate and approve the optical engine before their planned production periods. Completing qualification would clear an important manufacturing gate, but it would not guarantee shipment volumes, system launch dates or adoption across large AI clusters.

Several commercial details remain undisclosed, including the customer products behind the expected ramps, committed order volumes, manufacturing yields and the allocation of the new funding among design, testing, packaging and supply-chain partners. Ayar now has more capital to pursue production readiness; the next decisive evidence will be completed qualification, followed by identifiable customer programs and measurable manufacturing output.

Also read:

• A Gantt Chart Shows the Deadline—Its Logic Determines Whether You Can Trust It
• Six Billion People Are Online, but the Internet’s AI Future Is Less Equal
• Moglix Adds AI to B2B Procurement, but Physical Trade Still Sets the Margins