LUT-LLM: Efficient Large Language Model Inference with Memory-based Computations on FPGAs

TL;DR

LUT-LLM introduces a memory-based FPGA accelerator for large language models, leveraging table lookups and vector quantization to significantly improve inference speed and energy efficiency compared to GPUs.

Contribution

This work presents the first FPGA-based LLM inference method using memory-based computations with vector quantization, enabling scalable deployment of large models.

Findings

Achieves 1.10 to 3.29 times faster generation speed than GPUs.

Provides 3.05 to 6.60 times higher energy efficiency.

Reduces arithmetic operations by 4 times.

Abstract

The rapid development of large language models (LLM) has greatly enhanced everyday applications. While many FPGA-based accelerators, with flexibility for fine-grained data control, exhibit superior speed and energy efficiency compared to GPUs, recent GPU-specific optimizations have diminished this advantage. When limited to arithmetic-based computation, FPGAs often underperform GPUs due to their comparatively fewer computational resources. To address this challenge, we exploit a key advantage of FPGAs over GPUs: abundant distributed on-chip memory embedded among computational units. We believe that shifting LLM inference from arithmetic-based to memory-based computations through table lookups can improve the efficiency on FPGAs to compete with GPUs. However, existing methods are inefficient or unable to scale and deploy language models due to algorithm and architecture design limitations. This paper introduces \textbf{LUT-LLM}, the first FPGA accelerator that deploy 1B+ language model with memory-based computation, leveraging vector quantization. We construct a performance model, evaluate multiple quantization schemes, and identify activation-weight vector co-quantization as the most effective approach. To support this scheme, LUT-LLM features (1) bandwidth-aware parallel centroid search to reduce decoding latency, (2) efficient 2D table lookups, and (3) a spatial-temporal hybrid design to reduce data caching for a higher throughput table lookup. We develop a training recipe that converts existing models to support table lookups with high accuracy and prototype LUT-LLM for Qwen 3 1.7B model on the AMD V80 FPGA, reducing arithmetic operations by $4\times$ and achieving a $1.10\sim3.29\times$ faster generation speed and a $3.05\sim 6.60\times$ higher energy efficiency than GPUs.