Advantest Corporation, which is the global leader in the production of ATE (automatic test equipment) for semiconductors, and OpenLight, which is a company pioneering the technology of heterogeneous silicon photonics integration and custom Photonic Application-Specific Integrated Circuit (PASIC) design, have recently revealed a partnership aimed at developing high volume scalable testing solutions for silicon photonics products.
The partners plan to create a universal, end-to-end testing platform for the industry. It will help to overcome the bottlenecks associated with the mass manufacturing of next generation optical components and speed up the worldwide deployment of ultra-fast interconnection technologies in hyper-scale data centers and telecommunications infrastructures.
Electro-Optical Testing Bottleneck in AI Factories Overcome
Due to rapid scaling of deep learning models and HPC (high-performance computing) networks, conventional copper switches reach their physical limitations in terms of throughput capacity, cooling and distance. In order to eliminate this limitation, silicon photonics and co-packaging optical (CPO) architectures are employed by hardware architects. As a result, optical engines can be placed near high power compute ASICs providing necessary latency, bandwidth, and power efficiency for AI tasks.
Also Read: Nextpower to Acquire Zimmermann PV-Steel Group to Expand Utility-Scale Solar Infrastructure
However, moving light onto a silicon chip introduces significant manufacturing complications. Unlike standard electronics that only require electrical verification, photonic devices require simultaneous testing of electrical paths and physical laser beams. Aligning sub-micron optical fibers to chips at commercial volumes introduces severe manufacturing latency, driving up test costs and hindering market scale-up.
Advantest and OpenLight are joining forces to break this bottleneck. By pairing Advantest’s massive automatic test equipment experience with OpenLight’s deep PASIC engineering insight, the partnership will establish automated testing workflows optimized to manage:
Complex Multi-Channel Optical Alignment: Designing advanced mechanical hardware and software routines to align laser streams rapidly across automated wafer-level test loops.
Integrated Laser Heterogeneous Testing: Evaluating embedded indium phosphide (InP) lasers directly integrated on silicon substrates, ensuring high yield before final packaging.
Dynamic Parametric Calibration: Capturing precise, real-time metrics on wavelength accuracy, insertion loss, and thermal fluctuations under intensive operating conditions.
“We work closely with our foundry partners to design and manufacture innovative PASICs to support the advancement of AI technology,” said Adam Carter, CEO of OpenLight. “We are pleased to collaborate with Advantest as part of this ecosystem. Their leadership in test will allow us to offer sophisticated solutions that deliver customer value and address demands for high-volume manufacturing.”
Accelerating Time-to-Market for Optical Co-Packaging
The joint initiative focuses heavily on industrializing testbeds for complex “Optical Engines”-highly integrated modules combining lasers, modulators, and photodetectors into a single footprint. Advantest plans to expand its core automated test platform architectures to accommodate these custom electro-optical workflows seamlessly. The resulting standardized test protocols will allow silicon foundries, assembly factories, and device developers to transition prototypes to high-volume manufacturing (HVM) lines with minimal technical or regulatory risk.
“As we broaden our silicon photonics capabilities, we are pleased to partner with OpenLight to accelerate end-to-end semiconductor technology development,” said Doug Lefever, representative director and group CEO of Advantest Corporation. “As AI and high-performance computing continue to drive demand for greater bandwidth and energy efficiency, this partnership enables us to innovate test solutions that help customers scale production and help bring next-generation photonic technologies to market faster, meeting the needs of the AI era.”