As AI, cloud, and 5G drive exponential data growth, traditional copper interconnects hit physical limits, prompting a shift to optical solutions—especially linear driver optics and UCIe standards—that promise higher bandwidth, longer reach, and better energy efficiency. As AI, cloud, and 5G drive exponential data growth, traditional copper interconnects hit physical limits, prompting a shift to optical solutions—especially linear driver optics and UCIe standards—that promise higher bandwidth, longer reach, and better energy efficiency. With the advent of optical interconnects, there is a promising alternative that could reshape the landscape of electronic design. By using light to transmit data, optical interconnects offer significant advantages over copper, including higher bandwidth, lower latency, and reduced power. The transition from copper to 800G optical interconnect — and increasingly 1. 6T — architectures is how next-generation AI infrastructure scales. As AI infrastructure scales at an unprecedented rate. Several startups have recently developed fiber interconnects on standard chiplets. If you pack too many copper wires together, eventually you'll run out of space—if they don't melt together first. AI data centers are encountering similar limitations with the electronic interconnects that shuttle. While copper still dominates ultra-short reach connectivity within racks, and pluggable optics remain the workhorse of scale-out data center fabrics, the panelists agreed that CPO represents the future of high-performance interconnect—particularly for scale-up GPU clusters where traditional modules. Lasers, CPO and OCS will be everywhere (indium phosphide, silicon photonics, co-packaged optics, optical circuit switch). The crowds were huge and the enthusiasm intense.