Research output
Publications
Peer-reviewed publications by the Systems and AI Lab
2026
4Tracking Phenological Status and Ecological Interactions in a Hawaiian Cloud Forest Understory using Low-Cost Camera Traps and Visual Foundation Models
Workshop on Camera Traps, AI and Ecology2026
Semi-Automated Large-Scale Dataset Annotation using a Few-Shot Object Detection Pipeline
The Ohio State University2026
Quality-Stratified Low-Rank Adaptation for Fine-Tuning Large Language Models on Private Document Collections
The Ohio State University2026
ProbFuse: A Domain-Decomposed Multi-Kernel Gaussian Process for County-Level Agricultural Risk Assessment and Conservation Policy Evaluation in the Maumee River Basin
The Ohio State University2026
2025
4Cyberinfrastructure for machine learning applications in agriculture: experiences, analysis, and vision
Frontiers in Artificial Intelligence 7, 14960662025
Understanding and Characterizing Communication Characteristics for Distributed Transformer Models
IEEE Micro2025
OHIO: Enhancing RDMA Scalability in Alltoall with Optimized Communication Overlap
IEEE Micro2025
SmartWilds: Multimodal Wildlife Monitoring Dataset
NeurIPS 2025 Workshop on Imageomics2025
2024
20Effective and Efficient Offloading Designs for One-Sided Communication to SmartNICs
2024 IEEE 31st International Conference on High Performance Computing2024
HyperSack: Distributed Hyperparameter Optimization for Deep Learning using Resource-Aware Scheduling on Heterogeneous GPU Systems
2024 IEEE 31st International Conference on High Performance Computing2024
Design and Implementation of Kernel-based MPI Reduction Operations for Intel GPU s
2024 IEEE 31st International Conference on High Performance Computing2024
Using BlueField-3 SmartNICs to Offload Vector Operations in Krylov Subspace Methods
2024 IEEE 31st International Conference on High Performance Computing2024
Scaling Large Language Model Training on Frontier with Low-Bandwidth Partitioning
2024 IEEE 31st International Conference on High Performance Computing2024
OHIO: Improving RDMA network scalability in MPI_alltoall through optimized hierarchical and intra/inter-node communication overlap design
2024 IEEE Symposium on High-Performance Interconnects (HOTI)2024
Characterizing Communication in Distributed Parameter-Efficient Fine-Tuning for Large Language Models
2024 IEEE Symposium on High-Performance Interconnects (HOTI)2024
Demystifying the communication characteristics for distributed transformer models
2024 IEEE Symposium on High-Performance Interconnects (HOTI)2024
The case for co-designing model architectures with hardware
53rd International Conference on Parallel Processing2024
The Case for Co-Designing Model Architectures with Hardware
Oak Ridge National Laboratory (ORNL)2024
Omb-cxl: A micro-benchmark suite for evaluating mpi communication utilizing compute express link memory devices
PEARC 20242024
OMB-FPGA: A Microbenchmark Suite for FPGA-aware MPIs using OpenCL and SYCL
PEARC 20242024
Infer-HiRes: Accelerating Inference for High-Resolution Images with Quantization and Distributed Deep Learning
PEARC 20242024
Design and Implementation of an IPC-based Collective MPI Library for Intel GPUs
PEARC 20242024
PML-MPI: A Pre-Trained ML Framework for Efficient Collective Algorithm Selection in MPI
IPDPS 20242024
HINT: Designing Cache-Efficient MPI_Alltoall using Hybrid Memory Copy Ordering and Non-Temporal Instructions
IPDPS 20242024
Exploiting inter-layer expert affinity for accelerating mixture-of-experts model inference
IPDPS 20242024
Accelerating MPI AllReduce Communication with Efficient GPU-Based Compression Schemes on Modern GPU Clusters
ISC High Performance 20242024
Accelerating large language model training with hybrid gpu-based compression
2024 IEEE 24th International Symposium on Cluster, Cloud and Internet2024
Accelerating communication with multi-HCA aware collectives in MPI
Concurrency and Computation: Practice and Experience 36 (1)2024
2023
23Optimized All-to-All Connection Establishment for High-Performance MPI Libraries Over InfiniBand
2023 IEEE 30th International Conference on High Performance Computing2023
Flover: A Temporal Fusion Framework for Efficient Autoregressive Model Parallel Inference
2023 IEEE 30th International Conference on High Performance Computing2023
MPI4Spark Meets YARN: Enhancing MPI4Spark through YARN support for HPC
2023 IEEE International Conference on Big Data (BigData)2023
HARVEST: High-Performance Artificial Vision Framework for Expert Labeling using Semi-Supervised Training
2023 IEEE International Conference on Big Data (BigData)2023
Benchmarking Modern Databases for Storing and Profiling Very Large Scale HPC Communication Data
International Symposium on Benchmarking, Measuring and Optimization2023
Democratizing hpc access and use with knowledge graphs
SC'23 Workshops2023
Mpi-xccl: A portable mpi library over collective communication libraries for various accelerators
SC'23 Workshops2023
Designing and optimizing GPU-aware nonblocking MPI neighborhood collective communication for PETSc
IPDPS 20232023
Designing in-network computing aware reduction collectives in mpi
2023 IEEE Symposium on High-Performance Interconnects (HOTI)2023
Battle of the bluefields: An in-depth comparison of the bluefield-2 and bluefield-3 smartnics
2023 IEEE Symposium on High-Performance Interconnects (HOTI)2023
Dpu-bench: A micro-benchmark suite to measure offload efficiency of smartnics
PEARC 20232023
Optimizing amber for device-to-device gpu communication
PEARC 20232023
Enabling reconfigurable HPC through MPI-based inter-FPGA communication
Proceedings of the 37th International Conference on Supercomputing2023
In-depth evaluation of a lower-level direct-verbs API on InfiniBand-based clusters: Early experiences
IPDPS 20232023
Accelerating distributed deep learning training with compression assisted allgather and reduce-scatter communication
IPDPS 20232023
A novel framework for efficient offloading of communication operations to bluefield smartnics
IPDPS 20232023
Mcr-dl: Mix-and-match communication runtime for deep learning
IPDPS 20232023
Sai: Ai-enabled speech assistant interface for science gateways in hpc
International Conference on High Performance Computing2023
Performance characterization of using quantization for DNN inference on edge devices
2023 IEEE 7th International Conference on Fog and Edge Computing (ICFEC)2023
Scamp: Scalable meta-parallelism for deep learning search
2023 IEEE/ACM 23rd International Symposium on Cluster, Cloud and Internet2023
Implementing and optimizing a gpu-aware mpi library for intel gpus: Early experiences
2023 IEEE/ACM 23rd International Symposium on Cluster, Cloud and Internet2023
High performance mpi over the slingshot interconnect
Journal of Computer Science and Technology 38 (1)2023
WoundNet: A Domain-Adaptable Few-Shot Classification Framework for Wound Healing Assessment
2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI)2023
2022
15ACCDP: accelerated data-parallel distributed DNN training for modern GPU-based HPC clusters
HiPC 20222022
Designing efficient pipelined communication schemes using compression in MPI libraries
HiPC 20222022
Accelerating broadcast communication with gpu compression for deep learning workloads
HiPC 20222022
Efficient personalized and non-personalized alltoall communication for modern multi-HCA GPU-based clusters
HiPC 20222022
Spark meets mpi: Towards high-performance communication framework for spark using mpi
2022 IEEE International Conference on Cluster Computing (CLUSTER)2022
Designing hierarchical multi-HCA aware allgather in MPI
51st International Conference on Parallel Processing2022
Network assisted non-contiguous transfers for GPU-aware MPI libraries
2022 IEEE Symposium on High-Performance Interconnects (HOTI)2022
High performance mpi over the slingshot interconnect: Early experiences
PEARC 20222022
Arm meets Cloud: A Case Study of MPI Library Performance on AWS Arm-based HPC Cloud with Elastic Fabric Adapter
IPDPS 20222022
Towards Java-based HPC using the MVAPICH2 library: early experiences
IPDPS 20222022
Highly efficient alltoall and alltoallv communication algorithms for gpu systems
IPDPS 20222022
OMB-Py: Python micro-benchmarks for evaluating performance of MPI libraries on HPC systems
IPDPS 20222022
Hey CAI-Conversational AI Enabled User Interface for HPC Tools
International Conference on High Performance Computing2022
Hy-Fi: Hybrid Five-Dimensional Parallel DNN Training on High-Performance GPU Clusters
International Conference on High Performance Computing2022
Accelerating mpi all-to-all communication with online compression on modern gpu clusters
International Conference on High Performance Computing2022
2021
15The MVAPICH project: Transforming research into high-performance MPI library for HPC community
Journal of Computational Science 52, 1012082021
Optimizing distributed dnn training using cpus and bluefield-2 dpus
IEEE Micro 42 (2)2021
Layout-aware hardware-assisted designs for derived data types in mpi
HiPC 20212021
Towards architecture-aware hierarchical communication trees on modern hpc systems
HiPC 20212021
Distmile: a distributed multi-level framework for scalable graph embedding
HiPC 20212021
Large-message nonblocking mpi_iallgather and mpi ibcast offload via bluefield-2 dpu
HiPC 20212021
Accelerating CPU-based distributed DNN training on modern HPC clusters using bluefield-2 DPUs
2021 IEEE Symposium on High-Performance Interconnects (HOTI)2021
INAM: cross-stack profiling and analysis of communication in MPI-based applications
PEARC 20212021
Scaling single-image super-resolution training on modern hpc clusters: Early experiences
IPDPS 20212021
Bluesmpi: Efficient mpi non-blocking alltoall offloading designs on modern bluefield smart nics
International Conference on High Performance Computing2021
Designing a ROCm-aware MPI library for AMD GPUs: early experiences
International Conference on High Performance Computing2021
Designing high-performance mpi libraries with on-the-fly compression for modern gpu clusters
IPDPS 20212021
Super: Sub-graph parallelism for transformers
IPDPS 20212021
Adaptive and hierarchical large message all-to-all communication algorithms for large-scale dense gpu systems
2021 IEEE/ACM 21st International Symposium on Cluster, Cloud and Internet2021
Efficient MPI-based communication for GPU-accelerated Dask applications
2021 IEEE/ACM 21st International Symposium on Cluster, Cloud and Internet2021
2020
14FALCON-X: Zero-copy MPI derived datatype processing on modern CPU and GPU architectures
Journal of Parallel and Distributed Computing 144, 1-132020
Blink: Towards Efficient RDMA-based Communication Coroutines for Parallel Python Applications
HiPC 20202020
Scalable mpi collectives using sharp: Large scale performance evaluation on the tacc frontera system
ExaMPI 20202020
Mpi meets cloud: Case study with amazon ec2 and microsoft azure
Fourth Annual Workshop on Emerging Parallel and Distributed2020
Accelerating GPU-based Machine Learning in Python using MPI Library: A Case Study with MVAPICH2-GDR
Workshop on Machine Learning in HPC Environments2020
Gems: Gpu-enabled memory-aware model-parallelism system for distributed dnn training
SC202020
A hierarchical and load-aware design for large message neighborhood collectives
SC202020
Dynamic kernel fusion for bulk non-contiguous data transfer on GPU clusters
2020 IEEE International Conference on Cluster Computing (CLUSTER)2020
Accelerated real-time network monitoring and profiling at scale using OSU INAM
PEARC 20202020
Nv-group: link-efficient reduction for distributed deep learning on modern dense gpu systems
34th ACM International Conference on Supercomputing2020
Communication-aware hardware-assisted MPI overlap engine
International Conference on High Performance Computing2020
Analyzing and understanding the impact of interconnect performance on HPC, Big Data, and deep learning applications
IPDPS 20202020
Efficient training of semantic image segmentation on summit using horovod and mvapich2-gdr
IPDPS 20202020
EReinit: Scalable and efficient fault-tolerance for bulk-synchronous MPI applications
Concurrency and Computation: Practice and Experience 32 (3)2020
2019
15High-performance adaptive MPI derived datatype communication for modern Multi-GPU systems
HiPC 20192019
Designing a profiling and visualization tool for scalable and in-depth analysis of high-performance GPU clusters
HiPC 20192019
Leveraging network-level parallelism with multiple process-endpoints for mpi broadcast
Workshop on Emerging Parallel and Distributed2019
OMB-UM: Design, implementation, and evaluation of CUDA unified memory aware MPI benchmarks
PMBS 20192019
Hypar-flow: Exploiting mpi and keras for scalable hybrid-parallel dnn training using tensorflow
arXiv preprint2019
Communication profiling and characterization of deep-learning workloads
IEEE Micro 40 (1)2019
Performance characterization of dnn training using tensorflow and pytorch on modern clusters
2019 IEEE International Conference on Cluster Computing (CLUSTER)2019
Designing scalable and high-performance MPI libraries on Amazon elastic fabric adapter
2019 IEEE Symposium on High-Performance Interconnects (HOTI)2019
Optimized large-message broadcast for deep learning workloads: MPI, MPI+ NCCL, or NCCL2?
Parallel Computing 852019
Efficient design for MPI asynchronous progress without dedicated resources
Parallel Computing 852019
Falcon: Efficient designs for zero-copy mpi datatype processing on emerging architectures
IPDPS 20192019
Design and characterization of shared address space mpi collectives on modern architectures
CCGrid 20192019
Scalable distributed dnn training using tensorflow and cuda-aware mpi
CCGrid 20192019
High performance distributed deep learning: a beginner's guide
PPoPP 20192019
Instance segmentation in remote sensing imagery using deep convolutional neural networks
2019 International Conference on contemporary Computing and Informatics (IC3I)2019
2018
5Networking and communication challenges for post-exascale systems
Frontiers of Information Technology & Electronic Engineering 19 (10)2018
Multi-Threading and Lock-Free MPI RMA Based Graph Processing on KNL and POWER Architectures
EuroMPI 20182018
Efficient asynchronous communication progress for MPI without dedicated resources
EuroMPI 20182018
Salar: Scalable and adaptive designs for large message reduction collectives
2018 IEEE International Conference on Cluster Computing (CLUSTER)2018
Exploiting hardware multicast and GPUDirect RDMA for efficient broadcast
IEEE Transactions on Parallel and Distributed Systems 30 (3)2018
2017
5Scalable reduction collectives with data partitioning-based multi-leader design
SC172017
A scalable network-based performance analysis tool for MPI on large-scale HPC systems
2017 IEEE International Conference on Cluster Computing (CLUSTER)2017
Contention-aware kernel-assisted MPI collectives for multi-/many-core systems
2017 IEEE International Conference on Cluster Computing (CLUSTER)2017
Efficient and scalable multi-source streaming broadcast on GPU clusters for deep learning
ICPP 20172017
Designing dynamic and adaptive MPI point-to-point communication protocols
International Conference on High Performance Computing2017
2016
5System-level scalable checkpoint-restart for petascale computing
ICPADS 20162016
Designing mpi library with on-demand paging (odp) of infiniband
SC'162016
INAM2: InfiniBand network analysis and monitoring with MPI
International Conference on High Performance Computing2016
Exploiting maximal overlap for non-contiguous data movement processing
IPDPS 20162016
SHMEMPMI-Shared memory based PMI for improved performance and scalability
CCGrid 20162016