Comparing Llama-2 and GPT-3 LLMs for HPC kernels generation

TL;DR

This paper compares Llama-2 and GPT-3 models in generating high-performance computing kernels across various programming languages and models, revealing differences in reliability and optimization.

Contribution

It provides a comparative analysis of Llama-2 and GPT-3 for HPC kernel generation, highlighting Llama-2's competitive accuracy and optimization capabilities.

Findings

Llama-2 shows competitive or superior accuracy to GPT-3.

Copilot generates more reliable but less optimized code.

Llama-2 produces less reliable but more optimized code.

Abstract

We evaluate the use of the open-source Llama-2 model for generating well-known, high-performance computing kernels (e.g., AXPY, GEMV, GEMM) on different parallel programming models and languages (e.g., C++: OpenMP, OpenMP Offload, OpenACC, CUDA, HIP; Fortran: OpenMP, OpenMP Offload, OpenACC; Python: numpy, Numba, pyCUDA, cuPy; and Julia: Threads, CUDA.jl, AMDGPU.jl). We built upon our previous work that is based on the OpenAI Codex, which is a descendant of GPT-3, to generate similar kernels with simple prompts via GitHub Copilot. Our goal is to compare the accuracy of Llama-2 and our original GPT-3 baseline by using a similar metric. Llama-2 has a simplified model that shows competitive or even superior accuracy. We also report on the differences between these foundational large language models as generative AI continues to redefine human-computer interactions. Overall, Copilot generates codes that are more reliable but less optimized, whereas codes generated by Llama-2 are less reliable but more optimized when correct.