In EDP, 16-bit architecture is understood to be a processor architecture whose main processing width is 16 bits . Processors that use a 16-bit architecture are often referred to as "16-bit processors". Similarly, computer programs designed for such an architecture are given the 16-bit attribute (e.g. “16-bit operating system ” or “ Windows 16-bit”).
- The DDP-116 from 1965 from the series 16 of Honeywell is one of the first, his successors, especially the DDP-516 and H316 , the first commercialized 16-bit computer
- The PDP-11 was the Digital Equipment Corporation's first 16-bit computer
- The TMS9900 was the first 16-bit main processor from Texas Instruments . He was taken in 1975 to mass production and in 1979 in the home computers TI-99/4 and TI-99 / 4A installed, which are generally regarded as the first affordable 16-bit home computers.
- The WDC-65816/65802 processors from Western Design Center are the 16-bit successors of the 8-bit 6502 series from MOS Technology
- The Intel 8086 , 80186 and 80286 of the Intel 80x86 family were equipped with 16-bit registers, the base addresses were extended by segment addressing and the EMS / XMS standards, so that memory of up to 1 MB directly and up to 16 MB addressed via mapping. Starting with the 80386 , the family was expanded to a 32-bit architecture.
- The Motorola 68000 and 68010 processors of the 68000 family were designed with an external 16-bit data bus and an internal 16-bit ALU, but the programming model was based on 32-bit processing from the outset, which was used in later models Processor family has been fully implemented.
- The TI MSP430 is a 16-bit RISC microcontroller from Texas Instruments .
In simplified terms, 16 bits means that the processors are designed in such a way that 16 bits (i.e. 2 bytes ) can be processed simultaneously or during one cycle. This includes the external and internal design of data and address buses, the execution of the register set and the processing units. Furthermore, the instruction set usually contains an implementation of some of the common addressing types specially adapted to 16 bits .
In the C programming language , the number of bits is reflected in the size of the data types void * and int . This happened at a time when the first 32-bit systems already existed. To represent the fast processor operations, the Int type was set to 16 bits. However, since this was too short for many calculations, the long type was set to double, namely 32 bits. This is called IP16 for short.
The first widespread operating systems such as DOS worked with IP16 (small model), but segmentation was added here early on , so that 16 + 16-segmented pointers also existed, which can effectively only be stored in 32 bits. Similarly, this can be abbreviated as LP32, which was then adopted for the 32-bit architectures. Most Unix-like systems were considered to be compatible with 32-bit architectures at an early stage as the ILP32 type model, in which the Int type is also 32-bit, although programs in 16-bit architectures can run slower by half the word length.