- HyperTransport is named after its developer HyperTransport Consortium, an industry association formed by AMD, in addition to eight other technology companies: Alliance Semiconductor, Apple Computer, Broadcom Corp., Cisco Systems, Nvidia, PMC-Sierra, Sun Microsystems and Transmeta. The technology itself was introduced on April 2, 2001.
- HyperTransport is generally described as a bidirectional serial/parallel high-bandwidth, low-latency point-to-point link. This means that the technology quickly transmits a high level of frequencies between two endpoints, with the capability of going in opposite directions. In fact, it is designed to provide considerably more bandwidth than similar technologies. One of such technologies is a computer communications interface called the front-side bus (FSB). During the 1990s to around the mid 2000s, manufacturers used the FSB for data transfer between the CPU and the part of the chipset called the northbridge, which in turn communicates with other PC components such as the RAM and expansion slots.
- At the time of publication, there are five versions of HyperTransport: Version 1.0 in 2001, 1.1 in 2002, 2.0 in 2004, 3.0 in 2006 and 3.1 in 2008. Each version is more advanced than the one preceding it. For example, HyperTransport 1.0 has a maximum frequency of 800 megahertz and peak bidirectional aggregate bandwidth of 12.8 gigabytes per second. The latest version, HyperTransport 3.0, has a peak frequency of 3.2 gigahertz and maximum bidirectional aggregate bandwidth of 51.2 GBps.
- The most popular application of HyperTransport is as an FSB replacement. AMD replaces the FSB with the HyperTransport on some of its Athlon, Opteron, Phenom, Sempron and Turion processors for a faster mode of communication. Other HyperTransport uses include interconnection Non-Uniform Memory Access (NUMA) multiprocessor computers, and forwarding of data within routers and switches.
- Advertisement -
- Advertisement -