General-purpose computing on graphics processing units (hardware)

This article is about GPU hardware capable of running General-purpose programs. For Software that runs on modern GPUs, see General-purpose computing on graphics processing units (software).

GPGPUs are GPUs capable of running General-purpose programs. The extent to which "general-purpose" is defined as being "general" varies considerably: varying from Microprocessor-grade to fully capable of running Operating Systems such as GNU/Linux. Examples of the former have instruction subsets similar to the 8086: MIAOW had bit manipulation, bitwise operations, Vector floating-point and branch.^[1] whereas Larrabee was capable of running a full Linux OS.^[2]

List of GPGPUs

• Two RISC-V efforts in development include a GPGPU from Esperanto^[3][4] and a consortium led by Atif Zazar to leverage RISC-V.^[5][6][7] Esperanto offers a "direct" SDK to program individual cores, including the "accompanying vector/tensor unit" of each.^[8]
• For Larrabee, Tom Forsyth^[9] relates that it was capable of running a full Linux Operating System.^[10] The only specialist hardware (fixed-function) was texture sampling units.^[11]
• Nyuzi is a general-purpose processor with SIMD units.^[12] A custom port of LLVM was developed.^[13]
• MIAOW implemented a subset of Southern Islands^[14][15]
• Vortex GPU implemented RISC-V RV32IMAF and RV64IMAFD (a General-Purpose CPU) and "bolted on" a SIMT execution engine with a minimal RISC-V ISA Extension.^[16] Vortex GPU followed the same architecture as Larrabee by only providing specialist texture sampling hardware due to FPGA size limitations.^[17]

See also

• GPU – Specialized electronic circuit; graphics acceleratorPages displaying short descriptions of redirect targets
• Single instruction, multiple threads – Parallel Execution model which works simultaneously on arrays of several numbers
• Single instruction, multiple data – Type of parallel processing