Projects
Below are my (mostly personal) projects
hpcg-cpu
High Performance Conjugate Gradient (HPCG) benchmark for CPU, with many optimizations.
hpcg-sycl
A vendor agnostic GPU implementation of High Performance Conjugate Gradient (HPCG) benchmark written in SYCL.
llm-onnx.c
An llm runner using onnx models written in pure C, with no dependencies.
llm-onnx.hip
An llm runner using onnx models written in hip, with NVIDIA and AMD GPU support.