Ayar Labs has partnered with Taiwanese ASIC company AIchip to develop scalable AI infrastructure based on co-packaged optics (CPO) technology.

The collaboration will see Ayar Labs combine its CPO technology with AIchip’s advanced packaging expertize and TSMC’s advanced packaging and process technology, including COUPE, TSMC-SoIC, and advanced process nodes.

In a statement, the two companies said the partnership aims to “accelerate the production and adoption of optical engines” to address data bottlenecks and enable the development of new system architectures, ultimately supporting “hyperscaler demands for advanced AI accelerators and platforms that deliver enhanced performance, efficiency, and scalability.”

Additional details about the partnership and the planned CPO solutions for AI data centers will be shared in the coming weeks.

“Ayar Labs’ co-packaged optics technology unlocks the next era of AI infrastructure by removing the limitations of copper interconnects,” said Mark Wade, CEO and co-founder of Ayar Labs. “By combining our optical I/O innovation with Alchip’s deep expertise in advanced packaging, we’re building an ecosystem to accelerate the transition to power-efficient, high-performance AI systems.”

Johnny Shen, chairman and CEO of Alchip, added: “Current and future AI workloads require innovative and often collaborative advanced packaging design expertise and production-ready solutions. Alchip has proven to the market that it has the entire skill set to serve Tier 1 hyperscale customers.

“We’re working with Ayar Labs to bring leading-edge optical I/O technology to demanding high-performance next-gen designs, helping hyperscalers achieve new levels of data throughput and energy efficiency.”

Founded in 2015, California-based Ayar Labs develops optical interconnect solutions based on open standards, optimized for AI training and inference workloads. 

In April 2025, the company unveiled the world’s first UCIe (Universal Chiplet Interconnect Express) optical interconnect chiplet.

Designed to support AI workloads, Ayar Labs said the offering will eliminate bottlenecks, maximizing artificial intelligence (AI) infrastructure performance and efficiency whilst reducing latency and power consumption.