Cover Image

MultiPath Memory Access: Breaking Host-GPU Bandwidth Bottlenecks in LLM Services

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Tang, Lingfeng, Zhang, Daoping, Chen, Junjie, Huang, Peihao, Jin, Feng, Xu, Chengguang, Chen, Yuxin, Sun, Feiqiang, Chen, Guo
Format:	Preprint
Published:	2025
Subjects:	Distributed, Parallel, and Cluster Computing Networking and Internet Architecture Performance
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

On the Bandwidth Consumption of Blockchains
by: Lebedev, Andrei, et al.
Published: (2026)

TraDE: Network and Traffic-aware Adaptive Scheduling for Microservices Under Dynamics
by: Chen, Ming, et al.
Published: (2024)

JANUS: Resilient and Adaptive Data Transmission for Enabling Timely and Efficient Cross-Facility Scientific Workflows
by: Esaulov, Vladislav, et al.
Published: (2025)

Accelerator-as-a-Service in Public Clouds: An Intra-Host Traffic Management View for Performance Isolation in the Wild
by: Zhao, Jiechen, et al.
Published: (2024)

Local Rendezvous Hashing: Bounded Loads and Minimal Churn via Cache-Local Candidates
by: Guan, Yongjie
Published: (2025)

EvalNet: A Practical Toolchain for Generation and Analysis of Extreme-Scale Interconnects
by: Besta, Maciej, et al.
Published: (2021)

Performance and Stability of Barrier Mode Parallel Systems with Heterogeneous and Redundant Jobs
by: Walker, Brenton, et al.
Published: (2025)

PANDAS: Peer-to-peer, Adaptive Networking for Data Availability Sampling within Ethereum Consensus Timebounds
by: Pigaglio, Matthieu, et al.
Published: (2025)

Joint Network Slicing, Routing, and In-Network Computing for Energy-Efficient 6G
by: Sasan, Zeinab, et al.
Published: (2024)

FPsPIN: An FPGA-based Open-Hardware Research Platform for Processing in the Network
by: Schneider, Timo, et al.
Published: (2024)

Energy-Efficient and High-Performance Data Transfers with DRL Agents
by: Jamil, Hasibul, et al.
Published: (2025)

CASPER: Carbon-Aware Scheduling and Provisioning for Distributed Web Services
by: Souza, Abel, et al.
Published: (2024)

Proactive Service Assurance in 5G and B5G Networks: A Closed-Loop Algorithm for End-to-End Network Slicing
by: Tran, Nguyen Phuc, et al.
Published: (2024)

AFLL: Real-time Load Stabilization for MMO Game Servers Based on Circular Causality Learning
by: Kang, Shinsuk, et al.
Published: (2026)

GigaAPI for GPU Parallelization
by: Suvarna, M., et al.
Published: (2025)

Parallelizing a modern GPU simulator
by: Huerta, Rodrigo, et al.
Published: (2025)

Edge-First Language Model Inference: Models, Metrics, and Tradeoffs
by: Jang, SiYoung, et al.
Published: (2025)

LLAMP: Assessing Network Latency Tolerance of HPC Applications with Linear Programming
by: Shen, Siyuan, et al.
Published: (2024)

On Capacity and Delay of Wireless Networks with Node Failures
by: Li, Wei, et al.
Published: (2026)

A Case for CATS: A Conductor-driven Asymmetric Transport Scheme for Semantic Prioritization
by: Rizvi, Syed Muhammad Aqdas
Published: (2026)

Network-Aware Scheduling for Remote Gate Execution in Quantum Data Centers
by: Pouryousef, Shahrooz, et al.
Published: (2025)

Benchmarking Quantum Data Center Architectures: A Performance and Scalability Perspective
by: Pouryousef, Shahrooz, et al.
Published: (2026)

QECO: A QoE-Oriented Computation Offloading Algorithm based on Deep Reinforcement Learning for Mobile Edge Computing
by: Rahmaty, Iman, et al.
Published: (2023)

Generative AI on the Edge: Architecture and Performance Evaluation
by: Nezami, Zeinab, et al.
Published: (2024)

From Skew to Symmetry: Node-Interconnect Multi-Path Balancing with Execution-time Planning for Modern GPU Clusters
by: Yao, Jinghan, et al.
Published: (2026)

Ultra Ethernet's Design Principles and Architectural Innovations
by: Hoefler, Torsten, et al.
Published: (2025)

Temporal-Aware GPU Resource Allocation for Distributed LLM Inference via Reinforcement Learning
by: Du, Chengze, et al.
Published: (2025)

Compiler Support for Speculation in Decoupled Access/Execute Architectures
by: Szafarczyk, Robert, et al.
Published: (2025)

Optimizing Intra-Container Communication with Memory Protection Keys: A Novel Approach to Secure and Efficient Microservice Interaction
by: Yashu, Fnu, et al.
Published: (2025)

Characterizing and Optimizing LLM Inference Workloads on CPU-GPU Coupled Architectures
by: Vellaisamy, Prabhu, et al.
Published: (2025)

Multi-stage Flow Scheduling for LLM Serving
by: Sun, Yijun, et al.
Published: (2026)

PerLLM: Personalized Inference Scheduling with Edge-Cloud Collaboration for Diverse LLM Services
by: Yang, Zheming, et al.
Published: (2024)

FAST: An Efficient Scheduler for All-to-All GPU Communication
by: Lei, Yiran, et al.
Published: (2025)

InfiniteHBD: Building Datacenter-Scale High-Bandwidth Domain for LLM with Optical Circuit Switching Transceivers
by: Shou, Chenchen, et al.
Published: (2025)

PeerSync: Accelerating Containerized Service Delivery at the Network Edge
by: Deng, Yinuo, et al.
Published: (2025)

Dynamic Hierarchical Birkhoff-von Neumann Decomposition for All-to-All GPU Communication
by: Wu, Yen-Chieh, et al.
Published: (2026)

An Online Fragmentation-Aware GPU Scheduler for Multi-Tenant MIG-based Clouds
by: Zambianco, Marco, et al.
Published: (2025)

Can Asymmetric Tile Buffering Be Beneficial?
by: Wang, Chengyue, et al.
Published: (2025)

DualPath: Breaking the Storage Bandwidth Bottleneck in Agentic LLM Inference
by: Wu, Yongtong, et al.
Published: (2026)

ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression
by: Fan, Ruibo, et al.
Published: (2026)