SCALE documentation
SCALE is a GPGPU programming toolkit that allows CUDA applications to be natively compiled for AMD GPUs. SCALE does not require the CUDA program or its build system to be modified. Support for more GPU vendors and CUDA APIs is in development. To get started: SCALE has several key innovations compared to other cross-platform GPGPU solutions: We validate SCALE by compiling open-source CUDA projects and running their tests. The following open-source projects are currently part of our nightly automated tests and pass fully: The following GPU targets are supported, and are covered by our nightly tests: The following GPU targets have undergone ad-hoc manual testing and "seem to work": We are working on supporting the following GPUs: Contact us if you want us to expedite support for a particular AMD GPU architecture. SCALE consists of: Instead of providing a new way to write GPGPU software, SCALE allows programs written using the widely-popular CUDA language to be directly compiled for AMD GP
What is SCALE?# SCALE is a CUDA-compatible GPU programming toolkit targeting both NVIDIA and AMD GPUs. Instead of porting programs away from CUDA, SCALE offers a vendor-neutral platform that allows users to keep their existing codebases without compromising on performance or developer productivity. Any questions, suggestions, or bug reports? Join our community and talk with our dev team directly. What can SCALE do?# Examples of what SCALE can be used for: Compile existing CUDA codebases for AMD GPUs, with zero code changes. Inline asm? No problem. Provide valuable compiler…
related reading
- CUDA Programming Guide — CUDA Programming Guidedocs.nvidia.com
- How to Optimize a CUDA Matmul Kernel for cuBLAS-like Performance: a Worklogsiboehm.com
- README | GPU Glossarymodal.com
- CUDA C++ Programming Guide (Legacy) — CUDA C++ Programming Guidedocs.nvidia.com
- CUDA - Wikipediaen.wikipedia.org
- GitHub - adam-maj/tiny-gpu: A minimal GPU design in Verilog to learn how GPUs work from the ground up · GitHubgithub.com
- Inside NVIDIA GPUs: Anatomy of high performance matmul kernels - Aleksa Gordićaleksagordic.com
- AI Chip Architecturesjacobpeake.com
- What happens when you run a CUDA kernelfergusfinn.com
- CUTLASS: Fast Linear Algebra in CUDA C++ | NVIDIA Technical Blogdeveloper.nvidia.com
- Rust GPUrust-gpu.github.io
- GitHub - wafer-ai/gpu-perf-engineering-resources: A curated resource list for learning AI performance engineering, from GPU fundamentals to production inference.github.com