The competition between copper and optical interconnects in AI data centers is intensifying, with both technologies set to coexist rather than replace each other. As AI cluster sizes grow, connectivity has become a critical bottleneck, with copper interconnects maintaining dominance in short-distance, high-density scenarios due to cost and power efficiency. Meanwhile, optical interconnects are favored for long-distance, high-bandwidth applications, supporting the expansion of data centers into larger computing clusters.
NVIDIA and Broadcom are advancing Co-Packaged Optics (CPO) technology, which integrates optical engines directly onto chips, promising significant improvements in energy efficiency and signal integrity. However, CPO faces challenges in manufacturing and maintenance, delaying widespread adoption until after 2028. In the interim, Linear Pluggable Optics (LPO) may serve as a transitional solution, offering reduced power consumption while retaining modular design benefits. The shift towards CPO is expected to redistribute industry profits from traditional optical module manufacturers to chip design and advanced packaging firms.
Copper and Optical Interconnects to Coexist in AI Data Centers
Disclaimer: The content provided on Phemex News is for informational purposes only. We do not guarantee the quality, accuracy, or completeness of the information sourced from third-party articles. The content on this page does not constitute financial or investment advice. We strongly encourage you to conduct you own research and consult with a qualified financial advisor before making any investment decisions.
