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.

翻译：我们提出QUICK——一系列新颖的优化CUDA内核，用于高效推理量化大语言模型（LLM）。QUICK解决了当前最先进混合精度矩阵乘法内核中的共享内存bank冲突问题。该方法通过离线交错处理LLM的量化权重矩阵，跳过反量化后的共享内存写回操作。实验表明，在较大批次规模上，QUICK较AutoAWQ现有内核实现高达1.91倍加速；在代表性LLM模型上，于多种NVIDIA GPU设备中实现高达1.94倍的吞吐量提升。