EPD-Serve: A Flexible Multimodal EPD Disaggregation Inference Serving System On Ascend

• URL: http://arxiv.org/abs/2601.11590v1
• Date: Mon, 05 Jan 2026 03:17:15 GMT
• Title: EPD-Serve: A Flexible Multimodal EPD Disaggregation Inference Serving System On Ascend
• Authors: Fan Bai, Pai Peng, Zhengzhi Tang, Zhe Wang, Gong Chen, Xiang Lu, Yinuo Li, Huan Lin, Weizhe Lin, Yaoyuan Wang, Xiaosong Li,
• Abstract summary: We propose EPD-Serve, a stage-level disaggregated inference serving system for multimodal models.<n>EPD-Serve decouples the inference pipeline into independent Encode, Prefill, and Decode stages.<n>Under high-concurrency scenarios, EPD-Serve improves end-to-end throughput by 57.37-69.48% compared to PD-disaggregated deployment.
• Score: 19.427351311875718
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: With the widespread adoption of large multimodal models, efficient inference across text, image, audio, and video modalities has become critical. However, existing multimodal inference systems typically employ monolithic architectures that tightly couple the Encode, Prefill, and Decode stages on homogeneous hardware, neglecting the heterogeneous computational characteristics of each stage. This design leads to inefficient resource utilization and limited system throughput. To address these issues, we propose EPD-Serve, a stage-level disaggregated inference serving system for multimodal models. EPD-Serve decouples the inference pipeline into independent Encode, Prefill, and Decode stages, enabling logical isolation and flexible co-located deployment through dynamic orchestration. Leveraging the Ascend interconnect topology, EPD-Serve introduces asynchronous feature prefetching between Encode and Prefill stages and a hierarchical grouped KV cache transmission mechanism between Prefill and Decode stages to improve cross-node communication efficiency. In addition, EPD-Serve incorporates multi-route scheduling, instance-level load balancing, and multi-stage hardware co-location with spatial multiplexing to better support diverse multimodal workloads. Comprehensive experiments on multimodal understanding models demonstrate that, under high-concurrency scenarios, EPD-Serve improves end-to-end throughput by 57.37-69.48% compared to PD-disaggregated deployment, while satisfying strict SLO constraints, including TTFT below 2000 ms and TPOT below 50 ms. These results highlight the effectiveness of stage-level disaggregation for optimizing multimodal large model inference systems.

Related papers

• Beyond Real Weights: Hypercomplex Representations for Stable Quantization [6.708338010963415]
  Multimodal language models (MLLMs) require large parameter capacity to align high-dimensional visual features with linguistic representations.<n>We introduce a progressive re parameterization strategy that compresses these models by gradually replacing dense feed-forward network blocks.<n>A residual schedule, together with lightweight reconstruction and knowledge distillation losses, ensures that the PHM modules inherit the functional behavior of their dense counterparts during training.
  arXiv  Detail & Related papers  (2025-12-09T12:10:57Z)
• Adapformer: Adaptive Channel Management for Multivariate Time Series Forecasting [49.40321003932633]
  Adapformer is an advanced Transformer-based framework that merges the benefits of CI and CD methodologies through effective channel management.<n>Adapformer achieves superior performance over existing models, enhancing both predictive accuracy and computational efficiency.
  arXiv  Detail & Related papers  (2025-11-18T16:24:05Z)
• NExT-OMNI: Towards Any-to-Any Omnimodal Foundation Models with Discrete Flow Matching [64.10695425442164]
  We introduce NExT-OMNI, an open-source omnimodal foundation model that achieves unified modeling through discrete flow paradigms.<n>Trained on large-scale interleaved text, image, video, and audio data, NExT-OMNI delivers competitive performance on multimodal generation and understanding benchmarks.<n>To advance further research, we release training details, data protocols, and open-source both the code and model checkpoints.
  arXiv  Detail & Related papers  (2025-10-15T16:25:18Z)
• CollaPipe: Adaptive Segment-Optimized Pipeline Parallelism for Collaborative LLM Training in Heterogeneous Edge Networks [57.95170323315603]
  We introduce CollaPipe, a distributed learning framework that integrates collaborative pipeline parallelism with federated aggregation to support self-evolving networks.<n>In CollaPipe, the encoder part is adaptively partitioned into variable-sized segments and deployed across mobile devices for pipeline-parallel training, while the decoder is deployed on edge servers to handle generative tasks.<n>To enhance training efficiency, we formulate a joint optimization problem that adaptively allocates model segments, micro-batches, bandwidth, and transmission power.
  arXiv  Detail & Related papers  (2025-09-24T07:54:01Z)
• ElasticMM: Efficient Multimodal LLMs Serving with Elastic Multimodal Parallelism [9.93378263858092]
  Multimodal large language models (MLLMs) handle images, videos, and audio by incorporating feature extractors and projection modules.<n>Current tightly coupled serving architectures struggle to distinguish between mixed request types.<n>We introduce Elastic Multimodal Parallelism (EMP), a new serving paradigm that elastically adapts to resource heterogeneity.
  arXiv  Detail & Related papers  (2025-07-14T08:53:48Z)
• FindRec: Stein-Guided Entropic Flow for Multi-Modal Sequential Recommendation [57.577843653775]
  We propose textbfFindRec (textbfFlexible unified textbfinformation textbfdisentanglement for multi-modal sequential textbfRecommendation)<n>A Stein kernel-based Integrated Information Coordination Module (IICM) theoretically guarantees distribution consistency between multimodal features and ID streams.<n>A cross-modal expert routing mechanism that adaptively filters and combines multimodal features based on their contextual relevance.
  arXiv  Detail & Related papers  (2025-07-07T04:09:45Z)
• Task-Oriented Multimodal Token Transmission in Resource-Constrained Multiuser Networks [19.42660454288912]
  We propose a task-oriented multimodal token transmission scheme for efficient multimodal information fusion and utilization.<n>To improve the efficiency of token transmission, we design a two-stage training algotithm, including cross-modal alignment and task-oriented fine-tuning.<n>We jointly optimize bandwidth, power allocation, and token length across users by using an alternating optimization method.
  arXiv  Detail & Related papers  (2025-05-06T14:17:05Z)
• EPS-MoE: Expert Pipeline Scheduler for Cost-Efficient MoE Inference [49.94169109038806]
  This paper introduces EPS-MoE, a novel expert pipeline scheduler for MoE that surpasses the existing parallelism schemes.<n>Our results demonstrate at most 52.4% improvement in prefill throughput compared to existing parallel inference methods.
  arXiv  Detail & Related papers  (2024-10-16T05:17:49Z)
• STLLM-DF: A Spatial-Temporal Large Language Model with Diffusion for Enhanced Multi-Mode Traffic System Forecasting [32.943673568195315]
  We propose the Spatial-Temporal Large Language Model (STLLM-DF) to improve multi-task transportation prediction. The DDPM's robust denoising capabilities enable it to recover underlying data patterns from noisy inputs. We show that STLLM-DF consistently outperforms existing models, achieving an average reduction of 2.40% in MAE, 4.50% in RMSE, and 1.51% in MAPE.
  arXiv  Detail & Related papers  (2024-09-08T15:29:27Z)
• Improving Efficiency of Diffusion Models via Multi-Stage Framework and Tailored Multi-Decoder Architectures [12.703947839247693]
  Diffusion models, emerging as powerful deep generative tools, excel in various applications. However, their remarkable generative performance is hindered by slow training and sampling. This is due to the necessity of tracking extensive forward and reverse diffusion trajectories. We present a multi-stage framework inspired by our empirical findings to tackle these challenges.
  arXiv  Detail & Related papers  (2023-12-14T17:48:09Z)
• Proximal Policy Optimization-based Transmit Beamforming and Phase-shift Design in an IRS-aided ISAC System for the THz Band [90.45915557253385]
  IRS-aided integrated sensing and communications (ISAC) system operating in the terahertz (THz) band is proposed to maximize the system capacity. Transmit beamforming and phase-shift design are transformed into a universal optimization problem with ergodic constraints.
  arXiv  Detail & Related papers  (2022-03-21T09:15:18Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.