One of the motivating factors behind the creation of label switching is to make better use of the facilities provided by lower levels. The case of greatest interest in this regard is ATM, where other, previously- employed methods of allocating and using virtual circuits tend to be inefficient, particularly for handling multicast traffic. Conversely, some method of taking advantage of ATM features for service guarantees and path establishment is desirable. Various solutions have been proposed and implemented, including Ipsilon's IP Switching, Ascend's IP Navigator, Cisco's Tag Switching, IBM's ARIS, and Toshiba's Cell Switching Router.
The situation for Ethernet by contrast is relatively less explored. With the advent of Fast (100Mb) and especially Gigabit Ethernet technologies, it seems likely that Ethernet will play an increasing role in the broader network environment, so better handling of Ethernet-based pathways becomes important. While it is true that current Ethernet standards do not include the vast panoply of services defined for ATM, with modern bridging algorithms and the beginnings of quality-of-service provisions, there is still enough complexity that the interaction of IP-aware label switches with ordinary Ethernet switches offers both potential benefits and pitfalls.