IEEE TPDS

2,283 papers361 directions
IEEE Transactions on Parallel and Distributed Systems (TPDS) is a research venue that has published work in the fields of parallel processing and distributed computing since 1994. Within the machine learning and artificial intelligence community, the venue serves as a platform for research focused on the computational infrastructure and systems-level optimizations required for large-scale modeling. Its contributions often address the intersection of hardware architecture and algorithmic efficiency, covering topics such as memory system overhead, resource management, and the evolution of hardware-specific GPU kernels. Recent publications in the venue highlight a focus on scaling and optimizing deep learning workflows. This includes research into load balancing for Mixture-of-Experts (MoE) models, federated learning for resource-constrained environments, and collaborative inference in edge computing. The venue also features work on hardware-grounded energy profiling for sustainable deep learning and the development of specialized accelerators for graph convolution networks and sparse tensor decompositions. These studies aim to improve the performance of streaming inference and multi-dimensional search through techniques like hierarchical pruning and recursive statistical updates. The venue’s corpus includes contributions from researchers such as Keqin Li, Yunhao Liu, and Torsten Hoefler. By addressing issues like geographic bias in global machine learning and the automation of GPU kernel optimization, the venue facilitates the development of more efficient and accessible AI systems. Its role in the community is defined by its emphasis on the underlying systems that support machine learning, ranging from high-performance computing I/O analysis to the implementation of sparse kernels on modern tensor cores.

Papers by Year

2026 64
2025 119
2024 121
2023 130
2022 212
2021 165
2020 130
2019 102
2018 99
2017 117
2016 134
2015 182
2014 197
2013 124
2012 99
2011 82
2010 55
2009 58
2008 52
2007 49
2006 36
2005 31
2004 23
2003 28
2002 16
2001 19
2000 13
1999 14
1998 11
1997 19
1996 18
1995 14
1994 19
1993 17
1992 8
1991 2
1990 6

Top Research Directions

Research directions most represented by papers from this venue.

Hierarchical Model Aggregation for Multi-Tier Networks 5
Communication Efficient Distributed Training of Large Models 5
Decentralized Model Training via Federated Learning 5
Mitigating Cold Start Latency in Serverless Computing 4
Federated Learning Robustness to Statistical Data Heterogeneity 4
Dynamic Workload Redistribution in Parallel Systems 4
Hyperparameter Sensitivity in Federated Learning 4
Overcoming the Interconnect Bottleneck in Memory Offloading 4
Balancing Power Law Workloads in Parallel Graph Processing 4
Overcoming the Hardware Homogeneity Assumption 4
Decentralized Aggregation for Robust Federated Learning 4
Automated Tuning of Hardware Execution Schedules 3
Aligning Logical Communication with Physical Hardware Topologies 3
The Throughput Delay Tradeoff in Resource Allocation 3
Joint Communication and Computation Resource Allocation 3

Top Authors

AuthorPapers
Hai Jin60
Song Guo51
José Duato42
Xiang-Yang Li33
Baochun Li31
Bingsheng He29
Kenli Li27
Xiaohua Jia27
Viktor Prasanna25
Xueyan Tang23
Laurence Tianruo Yang22
Xue Liu22
Philip S. Yu19
Qiang He18
Yanmin Zhu18
Changjun Jiang17
Feng Zhang17
Ling Liu17
Lionel Ni17
Sanglu Lu17

Impactful Papers

Papers from this venue that shaped research clusters.

Foundational

Influential

Related Venues

Venues with the most similar research cluster distributions.

Papers

TitleYear
A Generalized Scheme for Mapping Parallel Algorithms1993
An Availability Model for MIN-Based Multiprocessors1993
A New Theory of Deadlock-Free Adaptive Routing in Wormhole Networks1993
A Parallel Hash Join Algorithm for Managing Data Skew1993
A Parallel Sort Merge Join Algorithm for Managing Data Skew1993
Cost and Time-Cost Effectiveness of Multiprocessing1993
Dependence Uniformization: A Loop Parallelization Technique1993
Latin Squares for Parallel Array Access1993
Multicast Communication in Multicomputer Networks1993
NETRA: A Hierarchical and Partitionable Architecture for Computer Vision Systems1993
Optimal Algorithms on the Pipelined Hypercube and Related Networks1993
Performance Analysis of Buffer Coherency Policies in a Multisystem Data Sharing Environment1993
Prediction-Based Dynamic Load-Sharing Heuristics1993
Replication Algorithms in a Remote Caching Architecture1993
The DASH Prototype: Logic Overhead and Performance1993
Trapezoid Self-Scheduling: A Practical Scheduling Scheme for Parallel Compilers1993
Using Processor-Cache Affinity Information in Shared-Memory Multiprocessor Scheduling1993
Guest Editors' Introduction1992
Implementation of Production Systems on Message-Passing Computers1992
Interleaving a Join Sequence with Semijoins in Distributed Query Processing1992
Multiprocessor Implementation of Digital Filtering Algorithms Using a Parallel Block Processing Method1992
On the Parallel Computation of the Algebraic Path Problem1992
Optimal Broadcasting on the Star Graph1992
Performance Evaluation of Circuit Switched Multistage Interconnection Networks Using a Hold Strategy1992
Reliable Distributed Sorting Through the Application-Oriented Fault Tolerance Paradigm1992
Automatic Generation of Self-Scheduling Programs1991
Using the Dual Path Property of Omega Networks to Obtain Conflict-Free Message Routing1991
Depth-First Search Approach for Fault-Tolerant Routing in Hypercube1990
Design, Analysis, and Simulation of I/O Architectures for Hypercube1990
Efficient Algorithms for List Ranking and for Solving Graph Problems on the Hypercube1990
Error Recovery in Shared Memory Multiprocessors Using Private Caches1990
Pipelined Data Parallel Algorithms-I: Concept and Modeling1990
Pipelined Data Parallel Algorithms-II: Design1990

« prev Page 46/46