In simplified terms, 8 bits means that the processors are designed in such a way that 8 bits (i.e. 1 byte ) can be processed simultaneously or during one cycle ( word length ). Register and address bus are often 16 bits wide, the memory area is partially expanded by memory mapping .
8-bit processors are still used most frequently by all processors, but not as main processors in computers, but primarily in the form of embedded systems as microcontrollers . Indeed, a modern computer system contains a number of 8-bit processors, e.g. B. in the mouse and keyboard, on the motherboard and in the monitors. The majority of USB devices are also based on 8-bit processors.
8-bit processors have the following properties:
- 8-bit data bus (can also be found in stripped-down 16-bit architectures such as the 8088 and 68008 )
- mostly 16-bit address bus (the 8008 only had 14 bit, there are still 16-bit CPUs with only 16-bit address bus, such as the Z8000 )
- 8-bit registers, which can often be combined to form 16-bit address registers (however, there are often restrictions when changing memory pages )
- internal 4-bit or 8-bit ALU (the Z80 e.g. had a fixed 4-bit ALU)
- Often individual, but no continuous 16-bit operations available (Z80 supports a few 16-bit operations, such as ADD / ADC / SBC and INC / DEC)
- mostly no hardware integer multiplication and division (but often extended in later microcontrollers)
- No floating point support (the connection of floating point coprocessors mostly came with the transition to 16-bit architectures)
- no separation between user and operating system mode
- relatively poor support of high-level languages - compilers , some the size of the stack is limited.
In contrast, the following changes are typical for 16-bit architectures, although not all of them have to be present:
- 16-bit data bus (but also 8-bit data bus for slimmed-down types)
- Address buses with 16 bit to 24 bit
- 16-bit or partially 32-bit registers that can be shared frequently.
- internal 16-bit ALU
- consistent 16-bit operations available
- Hardware integer multiplication and division
- Floating point support or coprocessor interface for floating point support
- Separation between user and operating system mode (not always available, sometimes MMU modules required)
- good support of high-level language compilers, flexible stack size.
Well-known processors with 8-bit architecture are the following processor or microcontroller architectures :
- from Intel , the 8008 - and 8080 models
- from MOS the 6502 series
- from Motorola the 6800 series.
- from Zilog the Z80 series
- from Intel and others the models of the MCS-51 and MCS-48 series (very widely used for embedded applications).
- from Microchip the PIC 10 to PIC18 series
- from Cypress the PSoC series
- from NEC the 78K0 series
- from Atmel the AVR family: completely redesign of an 8-bit RISC architecture
as well as various other designs.
Processors from the MCS-51 series are produced by many different manufacturers in very many variations, but in terms of quantity they are significantly behind the PIC processors from Microchip and the Motorola 6800 derivatives such as the 68HC05 and 68HC08 from Freescale , both of which make up more than a quarter of the market.