Difference between revisions of "Resource:Seminar"

From MobiNetS
Jump to: navigation, search
(wenliang updates seminars)
 
(255 intermediate revisions by 6 users not shown)
Line 1: Line 1:
{{SemNote
{{SemNote
|time='''2022-4-29 10:20'''
|time='''2026-04-10 10:30'''
|addr=4th Research Building A527-B
|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]].
}}
}}


===Latest===
===Latest===
{{Latest_seminar
{{Latest_seminar
|abstract = This paper presents EMU, a framework that enables the emulation, snipping, and multiplexing of LoRa chirps on commercial IoT devices equipped with low-power sub-GHz transceivers, including those supporting LoRa itself. Chirp snipping consists in artificially removing a sequence of chips and in putting the radio in low-power mode, which allows to reduce energy consumption while still communicating reliably. Chirp multiplexing exploits the gaps introduced by chirp snipping to transmit portions of another chirp on a separate channel, which allows to concurrently transmit two LoRa packets and to increase the throughput. We build EMU as a modular framework and implement support for off-the-shelf LoRa and non-LoRa transceivers. We then evaluate its performance by comparing the reliability, efficiency, and receiver sensitivity achieved by EMU with that of traditional LoRa for different physical layer settings. We finally showcase EMU’s ability to send packets over two channels simultaneously, thereby improving the uplink throughput of LoRaWAN, and demonstrate that even non-LoRa transceivers employing EMU can communicate to a LoRaWAN gateway, enabling new use cases and expanding the applicability of LoRa technology.
|abstract = To effectively utilize heterogeneous specialized hardware units in modern GPUs, such as TensorCores and Tensor Memory Accelerators, this paper introduces PipeThreader, a new DNN compiler. PipeThreader proposes shifting scheduling functionality from hardware to software so as to enable more efficient and sophisticated computation pipelining with minimal manual effort. This is achieved through sTask-graph, a new DNN computation abstraction, a hierarchical hardware abstraction that captures the capabilities of specialized units, and new scheduling primitives. As a result, PipeThreader can discover efficient pipeline scheduling for well-studied DNN architectures like FlashAttention, achieving comparable or even superior performance. Additionally, it can uncover novel pipeline schemes for emerging models like Mamba2, delivering significantly better performance compared to state-of-the-art hand-crafted implementations. The code is open-sourced at https://github.com/tile-ai/tilelang.
|confname= IPSN 2022
|confname =OSDI'25
|link=http://www.carloalbertoboano.com/documents/yang22emu.pdf
|link = https://www.usenix.org/conference/osdi25/presentation/cheng
|title= EMU: Increasing the Performance and Applicability of LoRa through Chirp Emulation, Snipping, and Multiplexing
|title= PipeThreader: Software-defined pipelining for efficient DNN execution
|speaker=Wenliang
|speaker=Junzhe
}}
|date=2026-4-9
{{Latest_seminar
|abstract = Containers, originally designed for cloud environments, are increasingly popular for provisioning workers outside the cloud, for example in mobile and edge computing. These settings, however, bring new challenges: high latency links, limited bandwidth, and resource-constrained workers. The result is longer provisioning times when deploying new workers or updating existing ones, much of it due to network traffic. Our analysis shows that current piecemeal approaches to reducing provisioning time are not always sufficient, and can even make things worse as round-trip times grow. Rather, we find that the very same layer-based structure that makes containers easy to develop and use also makes it more difficult to optimize deployment. Addressing this issue thus requires rethinking the container deployment pipeline as a whole. Based on our findings, we present Starlight: an accelerator for container provisioning. Starlight decouples provisioning from development by redesigning the container deployment protocol, filesystem, and image storage format. Our evaluation using 21 popular containers shows that, on average, Starlight deploys and starts containers 3.0x faster than the current state-of-the-art implementation while incurring no runtime overhead and little (5%) storage overhead. Finally, it is backwards compatible with existing workers and uses standard container registries.
|confname= NSDI 2022
|link=https://www.usenix.org/system/files/nsdi22-paper-chen_jun_lin.pdf
|title=Starlight: Fast Container Provisioning on the Edge and over the WAN
|speaker=Jiangshu
}}
}}


=== History ===
{{Resource:Previous_Seminars}}
{{Resource:Previous_Seminars}}

Latest revision as of 10:37, 10 April 2026

Time: 2026-04-10 10:30
Address: 4th Research Building A518
Useful links: 📚 Readling list; 📆 Schedules; 🧐 Previous seminars.

Latest

  1. [OSDI'25] PipeThreader: Software-defined pipelining for efficient DNN execution, Junzhe
    Abstract: To effectively utilize heterogeneous specialized hardware units in modern GPUs, such as TensorCores and Tensor Memory Accelerators, this paper introduces PipeThreader, a new DNN compiler. PipeThreader proposes shifting scheduling functionality from hardware to software so as to enable more efficient and sophisticated computation pipelining with minimal manual effort. This is achieved through sTask-graph, a new DNN computation abstraction, a hierarchical hardware abstraction that captures the capabilities of specialized units, and new scheduling primitives. As a result, PipeThreader can discover efficient pipeline scheduling for well-studied DNN architectures like FlashAttention, achieving comparable or even superior performance. Additionally, it can uncover novel pipeline schemes for emerging models like Mamba2, delivering significantly better performance compared to state-of-the-art hand-crafted implementations. The code is open-sourced at https://github.com/tile-ai/tilelang.

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=3511"