Pith. sign in

REVIEW 6 cited by

Efficient Multi-round LLM Inference over Disaggregated Serving

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2602.14516 v2 pith:2GKAZ4IY submitted 2026-02-16 cs.DC

Efficient Multi-round LLM Inference over Disaggregated Serving

classification cs.DC
keywords multi-roundworkloadsampdinferenceprefillservingattainmentdisaggregated
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

With the rapid evolution of Large Language Models (LLMs), multi-round workflows, such as autonomous agents and iterative retrieval, have become increasingly prevalent. However, this raises hurdles for serving LLMs under prefill-decode (PD) disaggregation, a widely adopted paradigm that separates the compute-bound prefill phase and memory-bound decode phase onto individual resources. Specifically, existing systems overlook the interleaved prefill-decode workload pattern in multi-round inference, leading to sub-optimal handling of the incremental prefill workloads and model deployment for the two phases. In this work, we present AMPD, a brand new disaggregated serving framework for multi-round LLM inference. The core of AMPD is to coordinate the prefill workloads based on real-time workloads by adaptively determining where to carry out these workloads and how they are scheduled, in order to maximize service level objective (SLO) attainment. In addition, we tailor a planning algorithm for our scenario, facilitating the deduction of optimal resource allocation and parallel strategies for the two phases. Empirical results demonstrate that AMPD substantially improves SLO attainment compared to state-of-the-art baselines.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. TurboServe: Serving Streaming Video Generation Efficiently and Economically

    cs.DC 2026-06 unverdicted novelty 7.0

    TurboServe introduces the first serving system for streaming video generation workloads, using migration-aware placement and load-driven autoscaling to cut worst-case latency by 37.5% and GPU cost by 37.2%.

  2. Observation, Not Prediction: Conversation-Level Disaggregated Scheduling for Agentic Serving

    cs.DC 2026-06 unverdicted novelty 7.0

    Conversation-level scheduling in ConServe observes first-turn input length and KV occupancy to route prefill once and pin decoders, cutting p95 time-to-first-effective-token by 51% and improving energy efficiency by 7...

  3. HexAGenT: Efficient Agentic LLM Serving via Workflow- and Heterogeneity-Aware Scheduling

    cs.DC 2026-05 unverdicted novelty 7.0

    HexAGenT reduces the SLO scale required for timely agentic LLM workflow completion by an average of 20.1% at 95% attainment and 33.0% at 99% attainment on heterogeneous A100/H100/H200 clusters.

  4. Autopoiesis: A Self-Evolving System Paradigm for LLM Serving Under Runtime Dynamics

    cs.DC 2026-04 unverdicted novelty 7.0

    Autopoiesis uses LLM-driven program synthesis to evolve serving policies online during deployment, delivering up to 53% and average 34% gains over prior LLM serving systems under runtime dynamics.

  5. DelAC: A Multi-agent Reinforcement Learning of Team-Symmetric Stochastic Games

    cs.MA 2026-05 unverdicted novelty 6.0

    Team-symmetric games always have team-symmetric Nash equilibria solvable via linear complementarity problems, and the DelAC actor-critic MARL algorithm outperforms existing methods in simulations.

  6. KAIROS: Stateful, Context-Aware Power-Efficient Agentic Inference Serving

    cs.DC 2026-04 unverdicted novelty 6.0

    KAIROS reduces power by 27% on average (up to 39.8%) for agentic AI inference by using long-lived context to jointly manage GPU frequency, concurrency, and request routing across instances.