Loading organizations...
§ Private Profile · San Jose, CA, USA
Develops silicon photonics optical I/O solutions for chip-to-chip connectivity, enabling high-bandwidth, low-latency data transfer for HPC and AI.
Based in Santa Clara, California, Ayar Labs develops silicon photonics technology and optical input/output solutions for high-bandwidth data transfer in advanced computing. The company manufactures optical chiplets and external light sources to replace traditional electrical copper wires, recently demonstrating a four-terabit-per-second bidirectional optical connection. Operating with an estimated workforce of 100 to 200 employees, the enterprise has secured over $220 million in total venture funding, including a $130 million Series C round and a $25 million Series C1 extension to fund commercial operations. This hardware targets artificial intelligence infrastructure and cloud data centers, supported by strategic equity investments and corporate partnerships with NVIDIA, AMD, Intel Capital, GlobalFoundries, and Hewlett Packard Pathfinder. Ayar Labs was established in 2015 by co-founders Alex Wright-Gladstein, Chen Sun, Mark Wade, Milos Popovic, Vladimir Stojanovic, and Rajeev Ram.
Ayar Labs has raised $880.0M across 8 funding rounds.
Ayar Labs has raised $880.0M in total across 8 funding rounds.
Ayar Labs has raised $880.0M across 8 funding rounds. Most recently, it raised $500.0M Series E in March 2026.
Ayar Labs has raised $880.0M in total across 8 funding rounds.
Ayar Labs's investors include Gabe Cahill, 1789 Capital, Alchip, AMD Ventures, ARK Invest, Insight Partners, MediaTek, NVIDIA, Qatar Investment Authority, Sequoia Global Equities, Jordan Katz, Shef Osborn.
Ayar Labs is a technology company developing optical I/O solutions, including the TeraPHY™ optical I/O chiplet and SuperNova™ multi-wavelength remote light source, to accelerate data movement in AI infrastructure.[1][5] It serves hyperscalers, AI system builders, and data center operators by solving data bottlenecks in large-scale AI training and inference, delivering 5-10x higher bandwidth, 4-8x better power efficiency, and 10x lower latency compared to traditional copper-based interconnects and pluggable optics.[1][3][6] The company has demonstrated strong growth momentum through $130M Series C funding in 2022, first volume shipments, partnerships with NVIDIA, HPE, Lockheed Martin, Intel, Alchip, and TSMC, and expansion including a 2025 Hsinchu, Taiwan office.[1][4]
Ayar Labs emerged from over a decade of research collaboration among scientists from MIT, UC Berkeley, and CU Boulder, focusing on optical interconnects to overcome limitations in copper wiring for chip-to-chip data transfer.[3] Founded by CEO and co-founder Mark Wade, the company secured Series A funding but pivoted in 2019 from a research-centric perception to a market-ready innovator through rebranding efforts emphasizing its leadership in photonics for AI.[2] Key milestones include 2020's $35M Series B, first working silicon on 45nm photonics, Intel FPGA demo for PIPES, and CW-WDM MSA establishment; 2022's Series C, volume shipments, and major partnerships; and 2025's Taiwan office opening.[1]
Ayar Labs rides the explosive growth of generative AI and trillion-parameter LLMs, where traditional copper interconnects fail to match compute scaling amid Moore's Law limits on data movement.[1][3][6] Timing is critical as AI workloads demand multi-rack scale-up architectures; optical I/O addresses power density, bandwidth, and latency barriers, enabling profitable inference and training at exascale.[1][4][6] Market forces like surging AI infrastructure needs from hyperscalers favor Ayar, with NVIDIA's investment signaling ecosystem momentum toward photonics-based designs.[1][5] It influences the landscape by defining CPO standards, fostering supply chain collaborations, and paving optical connectivity for next-gen HPC/AI clusters.[2][4]
Ayar Labs is poised to dominate optical interconnects as AI scale-up intensifies, with partnerships like Alchip/TSMC enabling mass production of CPO engines for hyperscale ASICs.[4] Trends in silicon photonics, energy-constrained data centers, and composable AI architectures will propel adoption, potentially expanding to edge HPC and beyond.[1][5][6] Its influence may evolve into industry-standard setter, unlocking "light waves ahead" for sustainable AI growth and redefining data center economics from the bottlenecks Ayar Labs first targeted.