# AI infrastructure connectivity and optical technology trends > Live situation record from CLSTR: https://clstr.news/situations/ai-infrastructure-connectivity-and-optical-technology-trends > Updated: 2026-08-13T22:03:56.000Z. Sources: 5. Developments: 2. The focus of AI infrastructure development is shifting from the procurement of GPUs and High Bandwidth Memory (HBM) toward the connectivity between components. As AI data centers expand, traditional copper-based electrical connections are facing limitations regarding power efficiency and signal attenuation. To address this, optical communication technology is emerging as a critical solution, involving a transition from long-distance rack-to-rack connections to Co-Packaged Optics (CPO) and the eventual optical conversion of GPU chip interfaces. Simultaneously, next-generation memory technologies like Compute Express Link (CXL) are being developed to manage the massive memory demands of Large Language Model (LLM) inference through ‘memory pooling’. This technological shift is being driven by surging data-center traffic, which has created acute shortages of high-speed optical components. In response, cloud builders are expanding their vendor pools to mitigate physical bottlenecks. This demand has resulted in significant revenue growth for networking and semiconductor firms, while photonics providers are developing new infrastructure, such as near-package optics (NPO) and interconnect platforms, to overcome bandwidth constraints at the chip perimeter. Recent developments indicate that the rapid expansion of AI is creating critical bottlenecks in both fiber-optic connectivity and power supply. While optical technologies like CPO are expected to reshape architectures—potentially moving from integrated to flexible, distributed structures—power availability is emerging as a decisive constraint. Projections suggest that data center electricity consumption could outpace total US power generation by the mid-2030s, making energy efficiency and source availability central to the scalability of AI infrastructure. ## Timeline ### 2026-08-13: AI demand drives optical component shortages and photonic innovation AI-driven demand is causing acute optical component shortages, forcing data centers to diversify suppliers and adopt new photonic technologies to overcome electrical and bandwidth bottlenecks. 2 sources. https://clstr.news/cluster/ai-demand-drives-optical-component-shortages-and-photonic-innovation ### 2026-08-13: AI infrastructure faces power and fiber-optic bottlenecks AI expansion is hitting critical infrastructure limits, specifically regarding power availability and the need for advanced fiber-optic connectivity to replace traditional copper wiring. 3 sources. https://clstr.news/cluster/ai-infrastructure-shifts-focus-toward-optical-connectivity-and-cxl-technology --- Cite as: AI infrastructure connectivity and optical technology trends. CLSTR, https://clstr.news/situations/ai-infrastructure-connectivity-and-optical-technology-trends