NW
SCALE™ optical module solution for co-packaged optics (CPO)
PhotonicsApplications

SCALE™ optical module solution for co-packaged optics (CPO)

By GLOBALFOUNDRIES

GF's SCALE solution, or Silicon photonics Co-packaged Advanced Light Engine solution, is the industry's first Optical Compute Interconnect Multi-Source Agreement (OCI MSA) capable platform, exceeding the requirements for the OCI MSA's optical interconnect specification for modern AI scale-up architectures. Built with GF's advanced silicon photonics technology, the SCALE CPO solution utilizes both coarse and dense wavelength-division multiplexing (CWDM, DWDM) for bi-directional data transmission over each optical fiber for significant improvements in bandwidth density and system scalability versus traditional copper interconnects. GF has already demonstrated 8λ and 16λ bi-directional DWDM natively on its platform, a fundamental technology milestone that uniquely positions GF to support the industry's shift to CPO and accelerate the adoption of optical scale-up interconnects. GF's SCALE CPO solution and silicon photonics technology offer an advanced portfolio of fully-qualified photonic devices, such as 50Gbps and 100Gbps micro-ring modulators, coupled ring resonators and integrated photodiodes. Additional features include through silicon vias (TSVs) for high-speed signaling and power delivery and copper pad pitches ranging from 110μm down to sub-45μm for 2.5D/3D stacking from organic substrates to silicon interposers, enabling customers to move quickly from design to volume production. The platform integrates electrical ICs on single-digit advanced nodes, enabling optimization between best-in-class compute and state-of-the-art optics without compromising performance. While GF offers multiple fiber-attach approaches, the SCALE solution leverages broadband detachable fibers with flat insertion loss over the CWDM spectrum to future-proof scaling from 4λ in each direction to 8λ and beyond, while still enabling serviceability and known-good-die testability for next-generation AI interconnects.