CodeGEMM: A Codebook-Centric Approach to Efficient GEMM in Quantized LLMs
This addresses efficiency bottlenecks in quantized LLM inference for deployment, though it is incremental as it builds on existing codebook-based methods.
The paper tackles the latency and cache pressure issues in codebook-based quantization for LLM inference by introducing CodeGEMM, a kernel that replaces dequantization with precomputed inner products, achieving speedups of 1.83x and 8.93x on Llama-3 models at 2-bit quantization with comparable accuracy.
Weight-only quantization is widely used to mitigate the memory-bound nature of LLM inference. Codebook-based methods extend this trend by achieving strong accuracy in the extremely low-bit regime (e.g., 2-bit). However, current kernels rely on dequantization, which repeatedly fetches centroids and reconstructs weights, incurring substantial latency and cache pressure. We present CodeGEMM, a codebook-centric GEMM kernel that replaces dequantization with precomputed inner products between centroids and activations stored in a lightweight Psumbook. At inference, code indices directly gather these partial sums, eliminating per-element lookups and reducing the on-chip footprint. The kernel supports the systematic exploration of latency-memory-accuracy trade-offs under a unified implementation. On Llama-3 models, CodeGEMM delivers 1.83x (8B) and 8.93x (70B) speedups in the 2-bit configuration compared to state-of-the-art codebook-based quantization at comparable accuracy and further improves computing efficiency and memory subsystem utilization.