Difference between revisions of "Resource:Seminar"

From MobiNetS
Jump to: navigation, search
 
(52 intermediate revisions by 3 users not shown)
Line 1: Line 1:
{{SemNote
{{SemNote
|time='''2024-10-25 10:30-12:00'''
|time='''2025-09-25 10:30'''
|addr=4th Research Building A533
|addr=4th Research Building A518
|note=Useful links: [[Resource:Reading_List|📚 Readling list]]; [[Resource:Seminar_schedules|📆 Schedules]]; [[Resource:Previous_Seminars|🧐 Previous seminars]].
|note=Useful links: [[Resource:Reading_List|📚 Readling list]]; [[Resource:Seminar_schedules|📆 Schedules]]; [[Resource:Previous_Seminars|🧐 Previous seminars]].
}}
}}
Line 8: Line 8:


{{Latest_seminar
{{Latest_seminar
|abstract = Sparsely-activated Mixture-of-Expert (MoE) layers have found practical applications in enlarging the model size of large-scale foundation models, with only a sub-linear increase in computation demands. Despite the wide adoption of hybrid parallel paradigms like model parallelism, expert parallelism, and expert-sharding parallelism (i.e., MP+EP+ESP) to support MoE model training on GPU clusters, the training efficiency is hindered by communication costs introduced by these parallel paradigms. To address this limitation, we propose Parm, a system that accelerates MP+EP+ESP training by designing two dedicated schedules for placing communication tasks. The proposed schedules eliminate redundant computations and communications and enable overlaps between intra-node and inter-node communications, ultimately reducing the overall training time. As the two schedules are not mutually exclusive, we provide comprehensive theoretical analyses and derive an automatic and accurate solution to determine which schedule should be applied in different scenarios. Experimental results on an 8-GPU server and a 32-GPU cluster demonstrate that Parm outperforms the state-of-the-art MoE training system, DeepSpeed-MoE, achieving 1.13× to 5.77× speedup on 1296 manually configured MoE layers and approximately 3× improvement on two real-world MoE models based on BERT and GPT-2.
|abstract = Distributed Edge Computing (DEC) has emerged as a novel paradigm, owing to its superior performance in communication latency, parallel computing efficiency, and energy consumption. With the surge of tasks in generative artificial intelligence, DEC faces higher demands for parallel computing efficiency. Scheduling multiple tasks for simultaneous processing, rather than one-by-one handling, could enhance parallel efficiency. Multiple tasks have multi-dependencies, i.e., sequence dependency, attribute similarity, and attribute correlation. Utilizing the bidirectional edges of traditional graphs to represent multi-dependencies can lead to an explosion in quantity. A hypergraph, with its hyperedges capable of connecting any number of vertices, can significantly solve the above problem. However, the multi-dependencies are rarely studied in the current research, posing the challenges, including incapable representing and unable capturing of multi-dependency hypergraph. In this work, we introduce a Joint communication and computation scheduling for hypErgraph Tasks in DEC, namely HypeJet, To effectively represent multi-dependencies, we employ hypergraph construction to represent task attributes and utilize hypergraph partitioning to clarify and refine task attribute correlations, enhancing parallel efficiency. In response to the challenge of capturing multi-dependencies, we employ a scheduling mechanism with the hypergraph neural network that efficiently acquires higher-order attribute correlated information among convolution matrices, providing enriched contextual information on multi-dependencies that supports decision-making in scheduling tasks. The evaluations using real-world traces demonstrate an 18.07% improvement in parallel efficiency of task scheduling.
|confname =INFOCOM‘24
|confname =INFOCOM'25
|link = https://ieeexplore.ieee.org/abstract/document/10621327
|link = https://ieeexplore.ieee.org/abstract/document/11044587
|title= Parm: Efficient Training of Large Sparsely-Activated Models with Dedicated Schedules
|title= HyperJet: Joint Communication and Computation Scheduling for Hypergraph Tasks in Distributed Edge Computing
|speaker=Mengqi
|speaker= Yi Zhou
|date=2024-11-1
|date=2025-9-26
}}{{Latest_seminar
|abstract = Localization of networked nodes is an essential problem in emerging applications, including first-responder navigation, automated manufacturing lines, vehicular and drone navigation, asset tracking, Internet of Things, and 5G communication networks. In this paper, we present Locate3D, a novel system for peer-to-peer node localization and orientation estimation in large networks. Unlike traditional range-only methods, Locate3D introduces angle-of-arrival (AoA) data as an added network topology constraint. The system solves three key challenges: it uses angles to reduce the number of measurements required by 4× and jointly uses range and angle data for location estimation. We develop a spanning-tree approach for fast location updates, and to ensure the output graphs are rigid and uniquely realizable, even in occluded or weakly connected areas. Locate3D cuts down latency by up to 75% without compromising accuracy, surpassing standard range-only solutions. It has a 0.86 meter median localization error for building-scale multi-floor networks (32 nodes, 0 anchors) and 12.09 meters for large-scale networks (100,000 nodes, 15 anchors).
|confname =NSDI'25
|link = https://www.usenix.org/conference/nsdi25/presentation/garg
|title= Large Network UWB Localization: Algorithms and Implementation
|speaker=Bangguo
|date=2025-9-26
}}
}}
{{Latest_seminar
|abstract = HD map is a key enabling technology towards fully autonomous driving. We propose VI-Map, the first system that leverages roadside infrastructure to enhance real-time HD mapping for autonomous driving. The core concept of VI-Map is to exploit the unique cumulative observations made by roadside infrastructure to build and maintain an accurate and current HD map. This HD map is then fused with on-vehicle HD maps in real time, resulting in a more comprehensive and up-to-date HD map. By extracting concise bird-eye-view features from infrastructure observations and utilizing vectorized map representations, VI-Map incurs low compute and communication overhead. We conducted end-to-end evaluations of VI-Map on a real-world testbed and a simulator. Experiment results show that VI-Map can construct decentimeter-level (up to 0.3 m) HD maps and achieve real-time (up to a delay of 42 ms) map fusion between driving vehicles and roadside infrastructure. This represents a significant improvement of 2.8× and 3× in map accuracy and coverage compared to the state-of-the-art online HD mapping approaches. A video demo of VI-Map on our real-world testbed is available at https://youtu.be/p2RO65R5Ezg.
|confname=Mobicom'23
|link = https://dl.acm.org/doi/abs/10.1145/3570361.3613280
|title= VI-Map: Infrastructure-Assisted Real-Time HD Mapping for Autonomous Driving
|speaker=Wangyang
|date=2024-11-1
}}
{{Resource:Previous_Seminars}}
{{Resource:Previous_Seminars}}

Latest revision as of 21:23, 25 September 2025

Time: 2025-09-25 10:30
Address: 4th Research Building A518
Useful links: 📚 Readling list; 📆 Schedules; 🧐 Previous seminars.

Latest

  1. [INFOCOM'25] HyperJet: Joint Communication and Computation Scheduling for Hypergraph Tasks in Distributed Edge Computing, Yi Zhou
    Abstract: Distributed Edge Computing (DEC) has emerged as a novel paradigm, owing to its superior performance in communication latency, parallel computing efficiency, and energy consumption. With the surge of tasks in generative artificial intelligence, DEC faces higher demands for parallel computing efficiency. Scheduling multiple tasks for simultaneous processing, rather than one-by-one handling, could enhance parallel efficiency. Multiple tasks have multi-dependencies, i.e., sequence dependency, attribute similarity, and attribute correlation. Utilizing the bidirectional edges of traditional graphs to represent multi-dependencies can lead to an explosion in quantity. A hypergraph, with its hyperedges capable of connecting any number of vertices, can significantly solve the above problem. However, the multi-dependencies are rarely studied in the current research, posing the challenges, including incapable representing and unable capturing of multi-dependency hypergraph. In this work, we introduce a Joint communication and computation scheduling for hypErgraph Tasks in DEC, namely HypeJet, To effectively represent multi-dependencies, we employ hypergraph construction to represent task attributes and utilize hypergraph partitioning to clarify and refine task attribute correlations, enhancing parallel efficiency. In response to the challenge of capturing multi-dependencies, we employ a scheduling mechanism with the hypergraph neural network that efficiently acquires higher-order attribute correlated information among convolution matrices, providing enriched contextual information on multi-dependencies that supports decision-making in scheduling tasks. The evaluations using real-world traces demonstrate an 18.07% improvement in parallel efficiency of task scheduling.
  2. [NSDI'25] Large Network UWB Localization: Algorithms and Implementation, Bangguo
    Abstract: Localization of networked nodes is an essential problem in emerging applications, including first-responder navigation, automated manufacturing lines, vehicular and drone navigation, asset tracking, Internet of Things, and 5G communication networks. In this paper, we present Locate3D, a novel system for peer-to-peer node localization and orientation estimation in large networks. Unlike traditional range-only methods, Locate3D introduces angle-of-arrival (AoA) data as an added network topology constraint. The system solves three key challenges: it uses angles to reduce the number of measurements required by 4× and jointly uses range and angle data for location estimation. We develop a spanning-tree approach for fast location updates, and to ensure the output graphs are rigid and uniquely realizable, even in occluded or weakly connected areas. Locate3D cuts down latency by up to 75% without compromising accuracy, surpassing standard range-only solutions. It has a 0.86 meter median localization error for building-scale multi-floor networks (32 nodes, 0 anchors) and 12.09 meters for large-scale networks (100,000 nodes, 15 anchors).

History

2024

2023

2022

2021

2020

  • [Topic] [ The path planning algorithm for multiple mobile edge servers in EdgeGO], Rong Cong, 2020-11-18

2019

2018

2017

Instructions

请使用Latest_seminar和Hist_seminar模板更新本页信息.

    • 修改时间和地点信息
    • 将当前latest seminar部分的code复制到这个页面中
    • 将{{Latest_seminar... 修改为 {{Hist_seminar...，并增加对应的日期信息|date=
    • 填入latest seminar各字段信息
    • link请务必不要留空，如果没有link则填本页地址 https://mobinets.org/index.php?title=Resource:Seminar
  • 格式说明
    • Latest_seminar:

{{Latest_seminar
|confname=
|link=
|title=
|speaker=
}}

    • Hist_seminar

{{Hist_seminar
|confname=
|link=
|title=
|speaker=
|date=
}}

Retrieved from "http://mobinets.cn/site/index.php?title=Resource:Seminar&oldid=3428"