Understanding Computer System Buses
Uploaded by
ScribdTranslationsUnderstanding Computer System Buses
Uploaded by
ScribdTranslationsSystem buses
In computer architecture, the bus (or channel) is a digital system that transfers data between
the components of a computer or between computers. It consists of cables or tracks in
a printed circuit, devices such as resistors and capacitors in addition to integrated circuits.
In the early electronic computers, all buses were of the parallel type, so that
communication between the parts of the computer was done through tapes or many tracks in
the printed circuit, in which each conductor has a fixed function and the connection is simple
requiring only input and output ports for each device.
The trend in recent years is towards the use of serial buses such as USB, Firewire for
communications with peripherals replacing parallel buses, including cases like the
microprocessor with the chipset on the motherboard. This despite the fact that the serial bus has a logic
complex (requiring more computing power than the parallel bus) in exchange for speeds and
greater efficacy.
There are various bus specifications that define a set of mechanical characteristics.
such as connectors, cables, and cards, in addition to electrical and signal protocols.
Operation
The function of the bus is to allow the logical connection between different subsystems of a system.
digital, sending data between devices of different orders: from within them
integrated circuits, up to complete digital equipment that is part of supercomputers.
Most buses are based on metal conductors through which transmission occurs.
electrical signals that are sent and received with the help of integrated circuits that have an interface
of the given bus and are responsible for managing the signals and delivering them as useful data. The signals
Digital signals that are transmitted are data, address, or control signals.
Buses define their capacity according to the maximum sending frequency and the width of the
data. Generally, these values are inversely proportional: if there is a high
frequency, the data bandwidth must be small. This is because of the interference between the
signals (crosstalk) and the difficulty of synchronizing them increase with frequency, such that a bus
with few signals it is less susceptible to those problems and can operate at high speed.
All computer buses have special functions such as interrupts and DMA.
that allow a peripheral device to access a CPU or memory using the minimum of
resources.
First generation
The early computers had 2 bus systems, one for memory and another for the
other devices. The CPU had to access two systems with instructions for each one,
different protocols and synchronizations.
The company DEC noticed that the use of two buses was unnecessary if the addresses were combined.
memory with the peripherals in a single memory space (mapping), so that the
architecture was simplified by saving manufacturing costs on mass-produced equipment, such as
they were the first minicomputers.
The first microcomputers were based on the connection of several printed circuit boards.
to a passive Backplane bus that served as the axis of the system. The PU card was connected to that bus.
which performs the functions of a referee for communications with the other device cards
connected; the cards included memory, diskette and disk controllers, adapters of
video. The CPU wrote or read the data pointing to the address of the desired device
in the unique address space allowing information to flow through the main bus.
Among the most well-known implementations are the S-100 bus and the ISA bus used in various
microcomputers of the 70s and 80s. In both, the bus was simply an extension of the bus
of the processor in such a way that it operated at the same frequency. For example, in systems with
Intel 80286 processor the ISA bus had a frequency of 6 or 8 megahertz depending on the
processor
Second generation
The fact that the bus was passive and used the CPU as control represented several
problems for the expansion and modernization of any system with that architecture. Furthermore
that the CPU used a considerable part of its power to control the bus.
Since processors started to operate at higher frequencies, it became necessary
hierarchize the buses according to their frequency: the concept of system bus was created (connection
between the processor and the RAM) and expansion buses, making the use of a chipset necessary.
The ISA bus used as a backplane in the original IBM PC transitioned from being a system bus to one of
expansion, leaving its arbitration to an integrated part of the chipset and implementing a bus to a
highest frequency to connect the memory with the processor.
On the other hand, the Nubus bus was independent from its creation, it had its own controller and
presented a standard interface to the rest of the system, allowing its inclusion in different
architectures. It was used in various devices, including some from Apple and was characterized by
to have a width of 32 bits and some Plug and Play (self-configuration) capabilities, which made it
very versatile and ahead of its time. Among other examples of these autonomous buses are the
AGP and the PCI bus.
Third generation
Third generation buses are characterized by having point-to-point connections, unlike
of the buses mentioned above in which clock signals are shared. This is achieved by reducing
strongly the number of connections that each device has using serial interfaces.
Then each device can negotiate the link characteristics at the beginning of the connection and in
some cases dynamically, just as in communication networks. Among the
Notable examples include PCI-Express buses, Infiniband, and HyperTransport.
AGP VIDEO CARD