Back to Search Start Over

QUICK: Quantization-aware Interleaving and Conflict-free Kernel for efficient LLM inference

Authors :
Kim, Taesu
Lee, Jongho
Ahn, Daehyun
Kim, Sarang
Choi, Jiwoong
Kim, Minkyu
Kim, Hyungjun
Publication Year :
2024

Abstract

We introduce QUICK, a group of novel optimized CUDA kernels for the efficient inference of quantized Large Language Models (LLMs). QUICK addresses the shared memory bank-conflict problem of state-of-the-art mixed precision matrix multiplication kernels. Our method interleaves the quantized weight matrices of LLMs offline to skip the shared memory write-back after the dequantization. We demonstrate up to 1.91x speedup over existing kernels of AutoAWQ on larger batches and up to 1.94x throughput gain on representative LLM models on various NVIDIA GPU devices.<br />Comment: 9 pages, 8 figures

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2402.10076
Document Type :
Working Paper