adam-maj/tiny-gpu resource

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

12868 stars · 1241 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

A minimal GPU implementation written in Verilog, designed to teach how GPUs work at the hardware level. It includes fully documented architecture and ISA docs, matrix addition/multiplication kernels, and kernel simulation with execution traces.

Use cases

• learn how GPUs work from the ground up
• understand SIMD parallelization in hardware
• study GPU memory architecture and bandwidth constraints
• simulate simple GPU kernels with execution traces
• explore GPGPU and ML accelerator design principles

When to choose

• you want a small, readable codebase for learning GPU hardware architecture
• you need documented Verilog examples of GPU cores, threads, and memory
• you are studying how modern hardware accelerators like TPUs work

When to avoid

• you need a production-grade or feature-complete GPU implementation
• you want to run real graphics workloads
• you need a synthesizable design for an FPGA project without modifications

Facets

learning-resource · maturity stable

simulation developer-tools hardware education gpu-computing cross-platform verilog gpu-architecture hardware-design simd systemverilog educational

Member repositories

For agents

markdown · JSON · MCP: product_card(name="adam-maj/tiny-gpu")