The main part, processing unit and devices, of a microcomputer is a system unit. A system unit includes a board called a motherboard that holds a microprocessor chip (or a CPU), memory chips, and expansion slots. Electronic circuitry is printed on the board and it connects between two main parts of a microcomputer, the microprocessor and primary storage and other parts. The system unit is housed within the system cabinet. A system unit includes the following parts:

A Motherboard or system board is the main printed, flat circuit board in an electronic device such as microcomputers. The board contains expansion slots (sockets) that accept additional boards (expansion Cards). In a microcomputer, the motherboard contains the microprocessor, the primary storage chips (or main memory cards), the buses, and all the chips used for controlling the peripherals.

A microprocessor is a processor whose elements are miniaturized into one or a few integrated circuits contained in a single silicon microchip. It executes instructions. In a microcomputer, the central processing unit (CPU) is held on a single microprocessor. In order to function as a processor, it requires a system clock, primary storage, and power supply.
Several important lines of microcomputers use some families of microprocessor chips. Intel and Motorola are the major companies that produce important microprocessors for IBM compatible and Macintosh computers.

Microprocessor Capacity

The capacity of a microprocessor chip is represented in word sizes. A word size is the number of bits (e.g., 8, 16, or 32 bits) that a computer (CPU) can process at a time.
If word has more bits, the computer (CPU) are more powerful and faster. For example, a 16-bit-word computer can access 2 bytes (1 byte = 8 bits) at a time, while a 32-bit-word computer can access 4 bytes at a time. Therefore, the 32-bit computer is faster than the 16-bit computer.

CISC and RISC Chips

• CISC Chips: CISC stands for Complex Instruction Set Computer. It is pronounced "sisk." CISC is a computer architecture that has large sets of instructions. CISC machines have several hundred instructions. Intel's Pentium-chip uses CISC design.

• RISC Chip: RISC stands for Reduced Instruction Set Computer. RISC is a computer that executes a few number of instructions. In RISC, most programs generally use only a few instructions. This architecture has a small number of instructions built into the circuits and if those basic instructions are made to execute faster, then RISC computers increase performance. Although RISC machines are only around 30% faster than their CISC machines, RISC chips are less expensive to produce. That is, RISC is less expensive per MIPS.

• Controversy: The proponents of RISC claim that its lower cost and speed are an extreme advantage, while the opponents claim its improvements are not great. The opponents believe that new machine languages and greater performance improvements are going to come. Besides, the reduced instructions make software (e.g., compilers) generate more code to do. Thus, the choice is not obvious.

Central Processing Unit (CPU)

The central processing unit (CPU) is the computing part of the computer that interprets and executes program instructions. It is also known as the processor. In a microcomputer, the CPU is contained on a single microprocessor chip within the system unit. The CPU has two parts: the control unit and the arithmetic-logic unit.

Additional storage units called registers within control unit and ALU help make processing more efficient.

• Control Unit: A control unit is the circuitry that locates, retrieves, interprets and executes each instruction in the central processing unit. The control unit directs electronic signals between primary storage and the ALU, and between the CPU and input/output devices.

• Atithmetic-Logic Unit (ALU): ALU is a high-speed circuit part in the CPU. The arithmetic-logic unit (ALU) performs arithmetic (math) operations, logic (comparison) operations and related operations. The ALU retrieves alphanumeric data from memory and then does actual calculating and comparing. It sends the results of the operation back to memory again.

Memory Chips

A memory chip is a chip that holds programs and data either temporarily or permanently. The major categories of memory chips are RAMs and ROMs.

RAM Chips

RAM stands for random-access memory. Random- access memory holds the data or instructions that the CPU is presently processing. The type of primary storage is RAM. That is, a collection of RAM chips builds primary storage.
Whenever a CPU writes data or instructions to RAM, it wipes out the previous contents of RAM, and when a CPU read data or instructions from RAM, it keeps their contents.

ROM Chips

ROM stands for read-only memory. A ROM chip is a memory chip that stores instructions and data permanently. Its contents are placed into the ROM chip at the time of manufacture and cannot be modified by the user. A CPU can read and retrieve the instructions and data from the ROM chip, but it cannot change the contents in ROM.
ROM chips usually contain special instructions for computer operations such as ROM BIOS. The variations on the ROM chip are the following:

• PROM (Programmable Read-Only Memory): A permanent storage device that becomes a read-only memory after it is written once by the customer rather than by the chip manufacturer. For example, a software producer can write instructions onto the PROM using special equipment.

• EPROM (Erasable Programmable Read-Only Memory): EPROM is a reusable PROM-chip that can be erased by a special ultraviolet light. EPROM holds its content until erased and new instructions can be written on it.

• EEPROM (Electrically Erasable Programmable Read-Only Memory): EEPROM-chip can be erased, either within a computer or externally, by electric power. The process usually requires more voltage than the common +5 volts used in logic circuits.

Primary Storage (Memory)

Primary storage (internal storage, main memory or memory) is the computer's working storage space that holds data, instructions for processing, and processed data (information) waiting to be sent to secondary storage. Physically, primary storage is a collection of RAM chips.
The contents are held in primary storage only temporarily. Capacity varies with different computers. Data or instructions are stored in primary storage locations called addresses.

The clock is a device that generates periodic, accurately spaced signals used for several purposes such as regulation of the operations of a processor or generation of interrupts. The clock circuit uses the fixed vibrations generated from a quartz crystal to deliver a steady stream of pulses to the processor. The system clock controls the speed of all the operations within a computer.
The clock speed is the internal speed of a computer. The clock speed is expressed in megahertzes (MHz). 33 MHz means 33 million cycles per second. A computer processor's speed is faster if it has higher clock speed. For example, a 100-Mhz processor is four times as fast internally as the same processor running at 25MHz.

Open/Closed architectures

• Open Architecture: This architecture is a system whose specifications are made public to encourage third-party vendors to develop add-on products for it. Most microcomputers adopt open architecture. They allow users to expand their systems using optional expansion boards.

• Closed Architecture: This is a system whose technical specifications are not made public. With a machine that has closed architecture, users cannot easily add new peripherals.

Expansion Slots

Expansion slots are receptacles inside a system unit that printed circuit boards (expansion boards) are plugged into. Computer buyers need to look at the number of expansion slots when they buy a computer, because the number of expansion slots decides future expansion. In microcomputers, the expansion slots are directly connected to the bus.

Expansion Boards

Expansion boards are also called expansion cards, controller cards, plug-in boards, adapter cards, or interface cards. Expansion boards are printed circuit boards that have many electronic components including chips. They are plugged into expansion slots.
Expansion boards are connected to peripherals through ports located on the edge of expansion boards. Expansion boards include memory expansion cards (e.g., SIMM), I/O controller cards (e.g., SCSI Card), video display card, sound cards, communications cards, etc.

A port is an external connecting socket on the outside the computer. This is a pathway into and out of the computer. A port lets users plug in outside peripherals, such as monitors, scanners and printers.

Serial Ports

Serial ports are external I/O connectors used to attach modems, scanners or other serial interface devices to the computer. The typical serial ports use a 9-pin DB-9 or a 25-pin DB-25 connector. Serial ports transmit bits one after another on a single communications line. Serial lines frequently are used to link equipment that is not located close by.

Parallel Ports

Parallel ports are external I/O connectors on a computer used to hook up printers or other parallel interface devices. The parallel port uses a DB-25 connector. This port transmits several bits simultaneously. Parallel lines move information faster than serial lines do.

A bus is a data pathway between several hardware components inside or outside a computer. It not only connects the parts of the CPU to each other, but also links the CPU with other important hardware. The other important hardware includes memory, a disk control unit, a terminal control unit, a printer control unit, and a communications control unit. The capacity of a bus is expressed as bits. A larger capacity bus is faster in data transfer. For example, a 32-bit bus is faster than an 8-bit bus.

Three Main Bus Architectures

• ISA (Industry Standard Architecture): ISA is pronounced i- suh. This is the original PC bus architecture. It includes the 8-bit (PC, XT) and 16-bit (AT) buses in IBM personal computer series and compatibles. Now, it refers specially to the 16-bit AT bus.

• MCA (Micro Channel Architecture): A 32-bit bus used in IBM P/S 2 series and other IBM models. This architecture allows multiprocessing that allows several processors to work simultaneously. Micro channel architecture is not compatible with PC bus architecture.

• EISA (Extended Industry Standard Architecture): EISA is pronounced eesa. This is a bus standard for PCs that extends the AT bus (the ISA bus) architecture to a 32-bit bus. This architecture also allows more than one CPU to share the bus. The purpose of EISA is to extend and amend the old ISA standard, so that all existing AT expansion boards can work with an EISA slot.

Local Buses

The performance of a microcomputer is often restrained by the relatively slow video cards and other peripherals, which cannot keep up with today's fast CPUs. A local bus reduces the performance gap between the high-speed microprocessors and slower hard disks, video boards and other peripherals.
There are two local-bus systems available today. Each bus hopes to boost microcomputer performance for I/O-intensive tasks. They are a VL-Bus and a PCI local bus.

• VL-Bus (VESA Local Bus): VESA specification was introduced by the VESA (Video Electronics Standards Association). VL-Bus added peripheral components and connectors to the existing motherboard's 486 local bus and was available first. Performance of the VL-Bus architecture declines sharply when supporting more than two devices, and the specification is currently limited to a 32-bit data path and 33-MHz operation. This design is vanishing.

• PCI (Peripheral Component Interconnect): A PCI chip set adds a 64-bit-wide bus between the microprocessor and peripherals to offer a 64-bit data path. This chip supports speeds of 66-MHz. PCI can transfer data either 32- or 64-bits at a time. This architecture is developed by Intel, Compaq, DEC, IBM and NCR. \PCI technology incorporates a managing layer to route and manage data for efficient handling of high-speed data transfers between the microprocessor and peripherals.
  Its design goals are to produce a low-cost, high-performance interface and support future generations of peripherals. PCI provides excellent compatibility, higher throughput and automatic configuration of peripheral cards. PCI also has features such as expandability and plug-and-play flexibility.

• Comparison: Both technologies employ a microprocessor's local bus instead of the system input/output bus to rapidly exchange data between the processor and peripherals.\The VESA design reached the market first and is less expensive than PCI, but PCI is technically superior. A VL-Bus usually supports only two or three local-bus peripherals, while PCI can support up to 10 local buses. PCI uses fewer bus lines than VL-Bus. This enables PCI to eventually cost less to manufacture. PCI is now dominating the market.