Inference Acceleration for Large Language Models on CPUs

TL;DR

This paper presents a CPU-based inference acceleration method for large language models, achieving significant throughput improvements and energy efficiency, enabling practical deployment of LLMs in real-world applications.

Contribution

The paper introduces a parallelized CPU inference approach with batching and multi-worker strategies, significantly enhancing throughput and reducing power consumption for large language models.

Findings

18-22x increase in tokens/sec throughput

4x additional improvement with 4 workers

48.9% reduction in power consumption

Abstract

In recent years, large language models have demonstrated remarkable performance across various natural language processing (NLP) tasks. However, deploying these models for real-world applications often requires efficient inference solutions to handle the computational demands. In this paper, we explore the utilization of CPUs for accelerating the inference of large language models. Specifically, we introduce a parallelized approach to enhance throughput by 1) Exploiting the parallel processing capabilities of modern CPU architectures, 2) Batching the inference request. Our evaluation shows the accelerated inference engine gives an 18-22x improvement in the generated token per sec. The improvement is more with longer sequence and larger models. In addition to this, we can also run multiple workers in the same machine with NUMA node isolation to further improvement in tokens/s. Table 2, we have received 4x additional improvement with 4 workers. This would also make Gen-AI based products and companies environment friendly, our estimates shows that CPU usage for Inference could reduce the power consumption of LLMs by 48.9% while providing production ready throughput and latency.