Yes. There are multicast addresses and broadcast addresses. Let’s look at multicast first.
Multicast (point-to-multipoint) is a communication pattern in which a source host sends a message to a group of destination hosts. Multicasting can be best compared to a television broadcast. Imagine your local PBS (Public Broadcasting Service) station. It broadcasts one copy of its programming on a certain frequency that can be picked up by any device that is tuned to that channel.
While a network interface must accept traffic destined for its own unicast address and the broadcast address, accepting multicast traffic is an individual decision. Its possible that some networked systems’ users may choose to participate in more than one multicast group.
Multicast addresses can be recognized by an odd value in their most significant byte. That occurs when the low value bit in the first byte of the Ethernet address field is set to 01 as shown above.
Some ways to use multicast include:
An investment firm might run real-time stock quotes on the desktops of all the brokers. If the ticker was run as a unicast application, it would eat up a serious chunk of the available bandwidth. A broadcast version on the other hand would create hardware interrupts on every computer on that subnet, not just the machines used by the brokers. However, by using multicasting, the stock quotes are delivered to the right users with a minimum of network overhead. Decrease of the network load is one big advantage of multicasting.
There are many other applications for this type of data transmission including audio and video teleconferencing, distance learning, and data transfer to a large number of hosts.
The set of hosts listening on a specific IP multicast address is called a host group. Members can receive traffic to their unicast address and to their group address. Host group membership is dynamic, and hosts can join and leave the group at any time. There are no limitations to the size or location of a host group.
RFC#1700 contains a list of multicast Ethernet addresses. An updated version of the same list is also available at http://www.iana.org/assignments/ethernet-numbers.
Broadcast packets go to every device on the local media segment. The broadcast addresses contain all binary 1s, which protocol analyzers display as hexadecimal FF FF FF FF FF FF.
Broadcast packets should be used sparingly. By definition, they must be received and processed by every device on the local network segment, which causes each device to stop what is was doing, pass the packet up to the higher level protocols, examine the contents of the packet to see if a response is needed, then resume it’s prior processing. While processing a single broadcast may not degrade performance of an individual device, high volumes of broadcast traffic significantly impact performance. By default, routers block these broadcasts rather than forwarding them, which keeps MAC broadcasts local.
In defense of the broadcast, it provides a good way to search a local network for a particular device even though the system has only recently joined the network, or to advertise available resources.
As a reminder, when you look at a frame header in a protocol analyzer, you will see the target or destination MAC address first, followed by the source MAC address. While the destination may be Unicast, Multicast, or Broadcast, the source should only be Unicast. If it’s other than Unicast, it has probably been artificially created.