The convergence of voice, data and video onto a single network combined with large global consumer and corporate appetite for internetworking is leading to the continuous installation and upgrade of communication infrastructure worldwide. However, the broad variety of networking protocols used in wide, metropolitan, local and storage area networks creates unnecessary network complexity. Great simplifications in network management in could be accomplished if high-bandwidth signals can be transported over standard copper wires.

The install-base of 1000BASE-T (Gigabit Ethernet over Copper) [1] has begun to grow significantly. This deployment has been fueled in part by its seamless backward compatibility with older BASE-T solutions and by the availability of cheap, low-power physical layer solutions. However, the IEEE 802.3 body has almost completed work on the successor to 1000BASE-T (10GBASE-T) [2]. This paper outlines some of the problems that VLSI designers are solving and provides some estimates on feasibility of future CMOS devices. We outline some of the challenges for 10GBASE-T and present some of the performance figures that a 10GBASE-T PHY will have to meet.