DBAIMar 29

Stream2LLM: Overlap Context Streaming and Prefill for Reduced TTFT

arXiv:2604.163958.6h-index: 3
Predicted impact top 38% in DB · last 90 daysOriginality Incremental advance
AI Analysis

For LLM serving systems, this work addresses the latency-quality tradeoff in multi-tenant deployments with dynamic context arrivals.

Stream2LLM reduces time-to-first-token (TTFT) in LLM inference by overlapping context retrieval with inference, achieving up to 11x TTFT improvement while maintaining throughput parity with non-streaming baselines.

Context retrieval systems for LLM inference face a critical challenge: high retrieval latency creates a fundamental tension between waiting for complete context (poor time-to-first-token) and proceeding without it (reduced quality). Recent work mitigates this via streaming--overlapping retrieval with inference--but prior systems focus on single-request settings and overlook challenges in multi-tenant deployments where concurrent requests contend for GPU memory and scheduling must adapt to dynamic context arrivals. We present STREAM2LLM, a system that extends vLLM to support streaming prompts with adaptive scheduling and preemption for two distinct retrieval patterns: append-mode (progressive context accumulation) and update-mode (iterative refinement with cache invalidation). STREAM2LLM decouples scheduling decisions from resource acquisition, enabling flexible preemption strategies guided by hardware-specific cost models, and uses cache invalidation based on longest common prefix matching to minimize redundant computation when prompts change dynamically. To evaluate STREAM2LLM, we collect and characterize two large-scale, real-world streaming workloads based on web crawling and approximate nearest neighbor search. Our evaluation demonstrates that streaming architecture delivers up to 11x TTFT improvements, with cost-aware scheduling providing critical benefits under memory pressure, while maintaining throughput parity with non-streaming baselines.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes