HyperTransportв„ў System Architecture

While HyperTransport was initially developed to address bandwidth and scalability problems associated with moving data through the I/O subsystems of desktops and servers, the networking extensions bring a number of enhancements which permit the advantages of HyperTransport technology to be extended to communications processing applications. There are some major differences in the requirements of host-centric systems such as desktops and servers and communications processing systems.

Communications Processing Is Often Less Vertical

In communications applications, there may be a number of processors or coprocessors located in various corners of the topology. The host processor may assume responsibility for configuration and control of coprocessors and interface devices, while the coprocessors perform specialized data processing tasks. Because of the distributed responsibility for control and data handling tasks , these systems tend to be much less host processor-centric.

As a result of decentralizing data processing in communications systems, information flow may be omni-directional as coprocessors initiate transactions targeting devices under their control. When switch components are added to the topology, elaborate multi-port configurations are possible.

Communications Processing Example

Figure 19-2 on page 449 illustrates an example of decentralized data processing in a communications processing system. This HyperTransport-based network switch system translates and routes data between an Ethernet Local Area Network (LAN) and a Wide Area Network (WAN).

Some things to note about Figure 19-2 on page 449:

• Unlike a desktop or server system, the host processor in this system is not involved in most of the data processing. Traffic moving between the local area network (LAN) and wide area network (WAN) pass through the corresponding interface and move into the HyperTransport fabric. At the switch, packets are passed downstream to the coprocessors for protocol look up and translation, as well as a security screening. Once translated, message packets are passed back up to the switch for routing to the other interface.

• In this application, the host processor might be used to configure devices at boot time, program the coprocessors as needed, handle errors, collect statistics, etc. Host control events are represented by the tag (1) in Figure 19-2.

• Note how useful direct peer-to-peer transfers are in this type of topology. The coprocessors can pass data and messages to each other without involving other links. This greatly enhances system concurrency and efficiency.