 | This articleneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "24-bit computing" – news ·newspapers ·books ·scholar ·JSTOR(October 2009) (Learn how and when to remove this message) |
| Computer architecture bit widths |
|---|
| Bit |
| Application |
| Binary floating-pointprecision |
| Decimal floating-pointprecision |
Incomputer architecture,24-bitintegers,memory addresses, or otherdata units are those that are 24bits (3 octets) wide. Also, 24-bitcentral processing unit (CPU) andarithmetic logic unit (ALU) architectures are those that are based onregisters,address buses, ordata buses of that size.
Notable24-bit machines include theCDC 924 – a 24-bit version of theCDC 1604,CDC lower 3000 series,SDS 930 andSDS 940, theICT 1900 series, theElliott 4100 series, and theDatacraft minicomputers/Harris H series.[1]
The term SWORD is sometimes used to describe a 24-bit data type with the S prefix referring tosesqui.[citation needed]
The range of unsigned integers that can be represented in 24 bits is 0 to 16,777,215 (FFFFFF16 inhexadecimal). The range of signed integers that can be represented in 24 bits is −8,388,608 to 8,388,607.
The IBMSystem/360, announced in 1964, was a popular computer system with 24-bit addressing and32-bit general registers and arithmetic. The early 1980s saw the first popular personal computers, including theIBM PC/AT with anIntel 80286 processor using 24-bit addressing and16-bit general registers and arithmetic, and theAppleMacintosh 128K with aMotorola 68000 processor featuring 24-bit addressing and 32-bit registers.
TheeZ80 is a microprocessor and microcontroller family, with 24-bit registers and therefore 24-bit linear addressing, that isbinary compatible with the8/16-bitZ80.[citation needed]
The65816 is a microprocessor and microcontroller family with 16-bit registers and 24-bitbank switched addressing. It is binary compatible with the8-bit6502.[2]
Several fixed-pointdigital signal processors have a 24-bit data bus, selected as the basic word length because it gave the system a reasonable precision for the processing audio (sound). In particular, theMotorola 56000 series has three parallel 24-bit databuses, one connected to eachmemory space: program memory, data memory X, and data memory Y.[3]
Engineering Research Associates (later merged intoUNIVAC) designed a series of 24-bitdrum memory machines including the Atlas, its commercial version theUNIVAC 1101, theATHENA computer, theUNIVAC 1824 guidance computer, etc. Those designers selected a 24-bit word length because the Earth is roughly 40 million feet in diameter, and anintercontinental ballistic missile guidance computer needs to do theEarth-centered inertial navigation calculations to an accuracy of a few feet.[4][failed verification]
OpenCL has a built-in intrinsic for multiplication (mul24()) with two 24-bit integers, returning a 32-bit result. It is typically much faster than a 32-bit multiplication.[5]