Skip to main content
Definition Website Computer and Technology

Reduced Instruction Set Computer

Subject: Computer and Technology noun /rɪˈduːst ɪnˈstrʌk.ʃən sɛt kəmˈpjuː.tər/ 1 definition

  1. Definition (noun)

    Entry #226 · Updated

    A Reduced Instruction Set Computer (RISC) is a processor architecture built around a small, highly regular set of simple instructions rather than complex, multi-operation commands. Using uniform, typically fixed-length instruction formats, generous general-purpose register files, and a load-store design—where only dedicated load and store instructions access memory—RISC keeps hardware decoding minimal so instructions execute smoothly in a pipeline. This approach shifts complexity from processor control logic to the compiler, contrasting with Complex Instruction Set Computer (CISC) designs by trading higher instruction counts for simpler silicon, predictable timing, and high energy efficiency.

    Common Forms

    • Classic Fixed-Width RISC: Early and minimalist designs (such as MIPS I and SPARC) that use uniform 32-bit instructions, simple in-order pipelines, and strictly register-to-register arithmetic.

    • Compressed-Encoding RISC: Space-efficient variants (such as ARM Thumb-2 and the RISC-V C extension) that add 16-bit instructions alongside 32-bit formats to shrink firmware size on memory-constrained microcontrollers.

    • Superscalar Out-of-Order RISC: High-performance implementations (such as Apple M-series, ARM Neoverse, and IBM POWER) that decode, schedule, and execute multiple instructions per clock cycle out of order for desktops and servers.

    • Modular Extensible RISC: Configurable architectures (most notably RISC-V) built on a minimal base integer instruction set that pairs with optional standardized extensions for floating-point math, atomics, or vector processing.

    Example

    When compiling total = a + b for a 64-bit RISC-V processor, the compiler emits explicit ld instructions to load a and b from memory into registers x5 and x6, an add instruction to compute the sum in x7, and an sd instruction to store the result back to RAM. Because each operation uses a simple, uniform instruction format and follows the load-store boundary, the processor's hardwired decoder can feed the pipeline directly without microcode translation.