mlsanityData to 7 Oct 2026

DualScale: Energy-Efficient Disaggregated LLM Serving via Phase-Aware Placement and DVFS

arXiv 2602.18755 ↗First posted 21 Feb 2026cs.DCcited by 4

On the map: Efficient LLM inference and serving → Serving systems and request scheduling

Abstract (arXiv)

Prefill/decode disaggregation is increasingly adopted in LLM serving to improve the latency-throughput tradeoff and meet strict TTFT and TPOT SLOs. However, LLM inference remains energy-hungry: autoscaling alone is too coarse-grained to track fast workload fluctuations, and applying fine-grained DVFS under disaggregation is complicated by phase-asymmetric dynamics and coupling between provisioning and frequency control. We present DualScale, a two-tier energy optimization framework for disaggregated LLM serving. DualScale jointly optimizes placement and DVFS across prefill and decode using predictive latency and power models. At coarse timescales, DualScale computes phase-aware placement and baseline frequencies that minimize energy while satisfying SLO constraints. At fine timescales, DualScale dynamically adapts GPU frequency per iteration using stage-specific control: model predictive control (MPC) for prefill to account for queue evolution and future TTFT impact, and lightweight slack-aware adaptation for decode to exploit its smoother, memory-bound dynamics. This hierarchical design enables coordinated control across timescales while preserving strict serving SLOs. Evaluation on a 16x H100 cluster serving Llama 3.3 70B with production-style traces shows that DualScale meets TTFT/TPOT SLOs while reducing energy by up to 39% in prefill and 48% in decode relative to DistServe.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in Serving systems and request scheduling, one of 12 directions in Efficient LLM inference and serving. Builds on 14 of its 37 references that are in the map. First posted on arXiv 21 Feb 2026.

Builds on

14 of 37 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024SLO-aware GPU Frequency Scaling for Energy Efficient LLM Inference Serving5 Aug
  2. 2024Aladdin: Joint Placement and Scaling for SLO-Aware LLM Serving11 May
  3. 2024LoongServe: Efficiently Serving Long-Context Large Language Models with Elastic Sequence Parallelism15 Apr
  4. 2024Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve4 Mar

Other references · 10

  1. 2023ShareGPT4V: Improving Large Multi-Modal Models with Better Captionscited by 1,000
  2. 2023Splitwise: Efficient generative LLM inference using phase splittingcited by 536
  3. 2022DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scalecited by 400
  4. 2023PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPUcited by 210
  5. 2023CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Servingcited by 199
  6. 2022Accelerating Distributed MoE Training and Inference with Linacited by 109
  7. 2024Parrot: Efficient Serving of LLM-based Applications with Semantic Variablecited by 96
  8. 2022Zeus: Understanding and Optimizing GPU Energy Consumption of DNN Trainingcited by 78
  9. 2024Towards Greener LLMs: Bringing Energy-Efficiency to the Forefront of LLM Inferencecited by 59
  10. 2023Reducing Energy Bloat in Large Model Trainingcited by 22

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Enabling Spatially Fine-Grained DVFS in Neural Processing Units for Energy-Efficient LLM Servingcited by 0
  2. 2026Energy-Efficient LLM Serving via Disaggregated Attention--FFN and Flexible Frequency Scalingcited by 0
  3. 2026Phase-Decoupled, Model-Calibrated Power Control for Disaggregated LLM Servingcited by 0
  4. 2026Analytical Power-Aware Provisioning for Prefill-Decode Disaggregated AI Inferencecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025DynaServe: Unified and Elastic Execution for Dynamic Disaggregated LLM Serving
  2. 2025Boosting LLM Serving through Spatial-Temporal GPU Resource Sharing
  3. 2026Energy-Aware Scheduling for Serverless LLM Serving on Shared GPUs
  4. 2025DuetServe: Harmonizing Prefill and Decode for LLM Serving via Adaptive GPU Multiplexing
  5. 2025Sandwich: Joint Configuration Search and Hot-Switching for Efficient CPU LLM Serving
  6. 2025EcoServe: Enabling Cost-effective LLM Serving with Proactive Intra- and Inter-Instance Orchestration
  7. 2025VoltanaLLM: Energy-Efficient and SLO-Aware Disaggregated LLM Serving via Adaptive Frequency Control and State-Space Routing
  8. 2026RAPID-Serve: Resource-efficient and Accelerated P/D Intra-GPU Disaggregation