64bit

In computing, 64-bit describes an architecture, data path, or instruction set in which the standard unit of data processed or addressed is 64 bits wide. Practically, this expands the size of integers, pointers, and some floating-point representations, and increases the theoretical address space beyond 32-bit limits. The term is used for processors, operating systems, and software that natively operate on 64-bit values.

Most modern 64-bit CPUs implement 64-bit general-purpose registers and an ALU width of 64 bits, and use a 64-bit address bus to access memory. While the first widely used 64-bit desktop architecture was x86-64 (also known as AMD64), other major 64-bit families include ARM64 (AArch64), SPARC-64, and PowerPC 64. The Itanium (IA-64) architecture was 64-bit but did not achieve broad market adoption.

64-bit software ecosystems typically require a 64-bit operating system. They can run 64-bit programs that benefit from larger address spaces and wider pointers. There are 32-bit compatibility modes in many systems (such as WOW64 on Windows and multi-arch support on Linux). Data models differ by platform, for example LP64 and LLP64, affecting how pointers and long integers are sized.

Performance considerations vary. 64-bit code can improve performance for memory-intensive workloads by providing more registers and larger address spaces, but pointers and data structures consume more memory, potentially increasing cache pressure and memory bandwidth demands. Porting and maintaining 64-bit software also involves toolchain, ABI, and binary-compatibility considerations.

Today, 64-bit computing is standard in desktop, mobile, and server environments, enabling more memory support, improved precision in computations, and broader software ecosystems.