Research Papers 论文研究 4h ago Updated 2h ago 更新于 2小时前 44

Request-Level Energy Attribution for Batched LLM Serving 批处理LLM服务中的请求级能量归因

Introduces JouleShare, a framework for request-level energy attribution in batched LLM serving, addressing the gap between aggregate GPU telemetry and per-request energy accounting needs Uses an offline harness with vLLM to establish ground-truth Shapley energy values by replaying request subsets and integrating GPU power telemetry Proposes JCalib, a lightweight calibration model that predicts Shapley shares from inexpensive request features for real-time serving use Token-proportional attributi 提出JouleShare框架,解决批处理LLM服务中请求级能量归因难题 通过离线Shapley能量计算建立请求级能量真实值基准,填补现有研究空白 JCalib轻量校准模型将归因误差从0.44-0.46降至0.12-0.18,优于在线可用的单独测量基线 证明令牌比例归因在批处理环境下不可靠,测量Shapley真实值可有效校准低成本请求特征

55
Hot 热度
72
Quality 质量
65
Impact 影响力

Analysis 深度分析

TL;DR

  • Introduces JouleShare, a framework for request-level energy attribution in batched LLM serving, addressing the gap between aggregate GPU telemetry and per-request energy accounting needs
  • Uses an offline harness with vLLM to establish ground-truth Shapley energy values by replaying request subsets and integrating GPU power telemetry
  • Proposes JCalib, a lightweight calibration model that predicts Shapley shares from inexpensive request features for real-time serving use
  • Token-proportional attribution deviates significantly from exact Shapley (0.440–0.458 normalized L1), but JCalib reduces error to 0.116 (static) and 0.177 (continuous batching)
  • Demonstrates that token-based attribution is unreliable under batched execution and that measured Shapley ground truth can effectively calibrate low-cost features for fairer energy allocation

Why It Matters

This work addresses a critical gap in sustainable AI infrastructure: as LLM serving scales, accurate per-request energy accounting becomes essential for chargeback models, sustainability reporting, and workload optimization. The findings challenge the common industry assumption that token count is a sufficient proxy for energy cost, which has direct implications for pricing models and carbon accounting in production LLM systems.

Technical Details

  • JouleShare framework: Combines an offline energy attribution harness with an online calibration model. The harness replays request subsets under vLLM with a reproducible protocol, integrates GPU power telemetry, and computes exact Shapley energy values for each request.
  • JCalib calibration model: A lightweight model trained on Shapley ground truth to predict per-request energy shares from cheap-to-compute features (e.g., input/output token counts, batch position), enabling real-time attribution during serving.
  • Benchmarking across 16 model/workload combinations: Evaluated on three data-center GPUs under both static and continuous batching strategies, providing robust empirical validation.
  • Quantitative results: Token-proportional attribution shows normalized L1 errors of 0.440 (static batching) and 0.458 (continuous batching) versus exact Shapley. JCalib reduces these to 0.116 and 0.177 respectively, outperforming a standalone-measurement baseline that is impractical for online use.
  • Sampled Shapley extension: Used to scale measurements to larger group sizes where exact Shapley becomes computationally prohibitive, confirming that the attribution gap persists and a single offline calibration remains the most accurate deployable approach.

Industry Insight

  • Cloud providers and LLM API operators should reconsider token-based pricing models, as they systematically misallocate energy costs under batched serving; adopting calibrated attribution like JCalib could enable fairer chargeback mechanisms.
  • Sustainability and carbon-accounting teams can leverage measured Shapley ground truth to validate and improve existing attribution rules, ensuring compliance with emerging ESG reporting standards.
  • The offline-harness-then-calibrate paradigm demonstrated here offers a practical template for other resource-attribution problems (e.g., memory, latency) where exact per-request accounting is infeasible online but valuable for analytics and billing.

TL;DR

  • 提出JouleShare框架,解决批处理LLM服务中请求级能量归因难题
  • 通过离线Shapley能量计算建立请求级能量真实值基准,填补现有研究空白
  • JCalib轻量校准模型将归因误差从0.44-0.46降至0.12-0.18,优于在线可用的单独测量基线
  • 证明令牌比例归因在批处理环境下不可靠,测量Shapley真实值可有效校准低成本请求特征

为什么值得看

批处理LLM服务已成为主流部署模式,但GPU功率遥测仅为聚合数据,无法满足可持续性报告、费用回收和工作负载分析对请求级能量核算的需求。该研究首次提供可测量的请求级能量真实值,并给出可在线部署的校准方案,对数据中心能耗管理和碳核算具有重要实践价值。

技术解析

  • JouleShare框架:包含离线工具与在线校准模型两部分。离线工具通过vLLM重放请求子集,集成GPU功率遥测,计算每个请求的精确Shapley能量作为ground truth
  • JCalib校准模型:轻量级模型学习从廉价请求特征预测Shapley份额,实现在线高效归因,同时保持精确的批处理能量效率
  • 实验验证:覆盖16个模型/工作负载组合,在三种数据中心GPU上复现。静态批处理下令牌归因误差0.440,连续批处理0.458;JCalib分别降至0.116和0.177
  • 采样Shapley扩展:将测量参考扩展到更大请求组,验证单一离线校准仍为最准确的可部署规则

行业启示

  • 数据中心和LLM服务提供商应摒弃简单的令牌比例分摊,采用基于Shapley的归因方法提升能耗核算的公平性和准确性
  • 该框架可直接应用于ESG报告和碳足迹核算,帮助企业在可持续性披露中获得更可靠的请求级能耗数据
  • 批处理优化与能量效率需协同设计,建议将JCalib类轻量校准模型集成到vLLM等推理引擎中,实现吞吐量与能耗公平性的双重优化

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

LLM 大模型 Inference 推理 Research 科学研究 GPU GPU