TeleRAG: Efficient Retrieval-Augmented Generation Inference with Lookahead Retrieval

TL;DR

TeleRAG introduces a lookahead retrieval mechanism that significantly reduces latency and increases throughput in retrieval-augmented generation systems, enabling faster and more memory-efficient LLM inference.

Contribution

The paper presents TeleRAG, a novel system that employs lookahead retrieval and advanced scheduling to optimize RAG inference with minimal GPU memory usage.

Findings

Achieves up to 1.53x latency reduction for single queries

Attains 1.83x higher throughput in batched inference

Demonstrates good scalability in throughput

Abstract

Retrieval-augmented generation (RAG) extends large language models (LLMs) with external data sources to enhance factual correctness and domain coverage. Modern RAG pipelines rely on large datastores, creating a significant system challenge: achieving high throughput and low latency is difficult, especially when GPU memory is limited. To address these challenges, we propose TeleRAG, an efficient inference system that reduces latency and improves throughput with minimal GPU memory requirements. The core innovation of TeleRAG is lookahead retrieval, a prefetching mechanism that predicts required data and transfers them from CPU to GPU in parallel with LLM generation. In addition, TeleRAG adopts a prefetching scheduler and a cache-aware scheduler to support efficient multi-GPU inference with minimal overhead. Evaluations show TeleRAG achieves up to a 1.53x average end-to-end latency reduction (single-query) and 1.83x higher average throughput (batched), as well as good scalability in throughput. This confirms the practical utility of TeleRAG for faster and more memory-efficient deployments of RAG applications.