How does UCIe on chiplets enable optical interconnects in data centers?
By Jeff Shepard, Connector Tips (March 13, 2024)
Chiplets enable heterogeneous integration of various process nodes and materials to maximize performance. UCIe is a new die-to-die interconnect standard for high-bandwidth, low-latency, power-efficient, and cost-effective connectivity between chiplets. UCIe is also the first specification to include an interface that is compatible with optical links.
Large computing systems needed to support high-performance computing (HPC) applications like artificial intelligence (AI) and machine learning (ML) are becoming increasingly difficult to build. In conventional architectures, the density of compute and memory resources is increasing, creating bandwidth bottlenecks and interconnect challenges. High hardware density also causes power and cooling challenges.
To read the full article, click here
Related Chiplet
- DPIQ Tx PICs
- IMDD Tx PICs
- Near-Packaged Optics (NPO) Chiplet Solution
- High Performance Droplet
- Interconnect Chiplet
Related News
- Ayar Labs Unveils World's First UCIe Optical Chiplet for AI Scale-Up Architectures
- InfiniLink Secures $10M Funding from MediaTek, Sukna Ventures, and Egypt Ventures to Develop Integrated Optical Engines for Next-Gen AI Data Centers
- Cadence Accelerates SoC, 3D-IC and Chiplet Design for AI Data Centers, Automotive and Connectivity in Collaboration with Samsung Foundry
- Sarcina pioneers next-generation UCIe-A/S chiplet interconnects for AI systems
Latest News
- CoAsia SEMI Commences Supply of 3D IC SoCs: Korea’s First Case, Positioning 3D IC as the Next HBM
- Eliyan Secures $50 Million in Strategic Investments from Leading Hyperscalers and AI Infrastructure Providers to Accelerate Scalable AI Systems
- Veeco and imec develop 300mm compatible process to enable integration of barium titanate on silicon photonics
- Lightmatter Introduces Guide Light Engine for AI, Featuring VLSP Technology
- Lightmatter and GUC Partner to Produce Co-Packaged Optics (CPO) Solutions for AI Hyperscalers