Bitness

Bitness is a characteristic of a computer's architecture that describes the width of the processor's data path, registers, and address space. The most common bitness values are 8, 16, 32, and 64 bits, though specialized or historic CPUs have other widths.

Higher bitness allows processing larger data values natively and expands the theoretical addressable memory. It also determines the size of pointers and integers used by software, which affects memory usage and the design of the program's interfaces. Yet bitness alone does not determine performance: throughput depends on clock speed, memory hierarchy, compiler optimizations, and workload characteristics.

In software terms, bitness governs compatibility among operating systems, drivers, and applications. A 64-bit operating system can run 64-bit programs and, on many platforms, 32-bit programs as well through compatibility layers; 32-bit systems are typically limited to smaller address spaces and older software. Compilers target a specific bitness, producing binaries with 32-bit or 64-bit ABIs, which define calling conventions, data sizes, and library interfaces. Some ecosystems support mixed configurations, such as running 32-bit libraries on a 64-bit system.

Popular 64-bit architectures include x86-64 and ARM64, which extend their 32-bit predecessors while maintaining backward compatibility in many cases. 8-, 16-, and 32-bit designs are still common in embedded systems and legacy devices.

Overall, bitness is a useful shorthand for describing the broad data width and addressability of a system, but it is only one of many factors that influence performance and software compatibility.