This article is from the Token Ring FAQ, by James Messer James@NetworkUptime.com with numerous contributions by others.
From a functional point of view switching is exactly the same as
bridging. However switches use specially designed hardware called
Application Specific Integrated Circuits (ASICs) to perform the
bridging and packet-forwarding functionality (as supposed to
implementations using a central CPU and special software).
Consequently, switches are much faster than ancient bridges.
When you compare token-ring switches to multiport (token-ring)
bridges in more detail you can find more differences. For example
switches forward packets directly and at wire-speed from port x to
port y. However ancient multiport bridges are often implemented
using a internal token-ring segment. Consequently a packet being
source-routed from port x to port y makes two (!) hops (from the
segment attached to port x to the internal ring and from the
internal ring to the segment attached to port y). Please note that
there is a maximum on the number of hops a packet is allowed to make
(8 or 16, don't remember) and that the maximum aggregate throughput
of the multiport bridge is limited by the capacity of the internal
Other goodies token-ring switches often offer are support for
virtual LAN's and full duplex connections.