OHIO: Enhancing RDMA Scalability in Alltoall With Optimized Communication Overlap.

Bharath Ramesh 0005 Ohio State University, Columbus, OH, USA https://orcid.org/0000-0002-6430-8587 https://www.wikidata.org/entity/Q130808905 Bharath Ramesh Bharath Ramesh 0001 Western Sydney University, NSW, Australia National University of Singapore, Singapore https://sites.google.com/view/bharath-ramesh/ https://www.westernsydney.edu.au/icns/about/people/researchers/dr_bharath_ramesh https://scholar.google.com/citations?user=dh9MPA0AAAAJ https://ieeexplore.ieee.org/author/37085524942 https://au.linkedin.com/in/bharathrb https://orcid.org/0000-0001-8230-3803 https://www.wikidata.org/entity/Q88593030 Bharath Ramesh 0002 R.V. College of Engineering, Bengaluru, India Bharath Ramesh 0003 The University of New South Wales, Australia Bharath Ramesh 0004 Virginia Tech, Blacksburg, VA, USA Bharath Ramesh 0006 Polytechnic of Porto, ISEP - School of Engineering, Portugal

Tu Tran Goutham Kalikrishna Reddy Kuncham Bharath Ramesh 0005 Shulei Xu Hari Subramoni Dhabaleswar K. Panda 0001 OHIO: Enhancing RDMA Scalability in Alltoall With Optimized Communication Overlap. 36-45 2025 March - April 45 IEEE Micro 2 https://doi.org/10.1109/MM.2024.3524891 db/journals/micro/micro45.html#TranKRXSP25 streams/journals/micro

Kaushik Kandadi Suresh Bharath Ramesh 0005 Goutham Kalikrishna Reddy Kuncham Hari Subramoni Dhabaleswar K. Panda 0001 Towards Dynamic Message Passing Protocols for Stencil-Based Communication Patterns. 1-12 2025 CLUSTER https://doi.org/10.1109/CLUSTER59342.2025.11186470 conf/cluster/2025 db/conf/cluster/cluster2025.html#SureshRKSP25 streams/conf/cluster Nicholas Contini Jake Queiser Bharath Ramesh 0005 Hari Subramoni Dhabaleswar K. Panda 0001 A Streaming Collectives Interface Targeting Dataflow Acceleration and HPC Workloads. 1935-1950 2025 SC https://doi.org/10.1145/3712285.3759811 conf/sc/2025 db/conf/sc/sc2025.html#ContiniQ0S025

Tu Tran Bharath Ramesh 0005 Benjamin Michalowicz Mustafa Abduljabbar Hari Subramoni Aamir Shafi Dhabaleswar K. Panda 0001 Accelerating communication with multi-HCA aware collectives in MPI. 2024 36 Concurr. Comput. Pract. Exp. 1 https://doi.org/10.1002/cpe.7879 db/journals/concurrency/concurrency36.html#TranRMASSP24

Hooyoung Ahn Seonyoung Kim Yoo-Mi Park Woojong Han Shinyoung Ahn Tu Tran Bharath Ramesh 0005 Hari Subramoni Dhabaleswar K. Panda 0001 MPI Allgather Utilizing CXL Shared Memory Pool in Multi-Node Computing Systems. 332-337 2024 IEEE Big Data https://doi.org/10.1109/BigData62323.2024.10825804 conf/bigdataconf/2024 db/conf/bigdataconf/bigdataconf2024.html#AhnKPHAT0S024 Nawras Alnaasan Bharath Ramesh 0005 Jinghan Yao Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 HyperSack: Distributed Hyperparameter Optimization for Deep Learning using Resource-Aware Scheduling on Heterogeneous GPU Systems. 100-110 2024 HiPC https://doi.org/10.1109/HiPC62374.2024.00019 conf/hipc/2024 db/conf/hipc/hipc2024.html#Alnaasan0YSS024 Kaushik Kandadi Suresh Benjamin Michalowicz Nick Contini Bharath Ramesh 0005 Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 Using BlueField-3 SmartNICs to Offload Vector Operations in Krylov Subspace Methods. 155-165 2024 HiPC https://doi.org/10.1109/HiPC62374.2024.00029 conf/hipc/2024 db/conf/hipc/hipc2024.html#SureshMC0ASS024 Tu Tran Goutham Kalikrishna Reddy Kuncham Bharath Ramesh 0005 Shulei Xu Hari Subramoni Mustafa Abduljabbar Dhabaleswar K. Panda 0001 OHIO: Improving RDMA Network Scalability in MPI_Alltoall Through Optimized Hierarchical and Intra/Inter-Node Communication Overlap Design. 47-56 2024 HOTI https://doi.org/10.1109/HOTI63208.2024.00019 conf/hoti/2024 db/conf/hoti/hoti2024.html#TranKRXSAP24 streams/conf/hoti Hooyoung Ahn Seonyoung Kim Yoo-Mi Park Woojong Han Nick Contini Bharath Ramesh 0005 Mustafa Abduljabbar Dhabaleswar K. Panda 0001 Towards Accelerating k-NN with MPI and Near-Memory Processing. 608-615 2024 IPDPS (Workshops) https://doi.org/10.1109/IPDPSW63119.2024.00119 conf/ipps/2024w db/conf/ipps/ipdps2024w.html#AhnKPHC0A024 Bharath Ramesh 0005 Nick Contini Nawras Alnaasan Kaushik Kandadi Suresh Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. D. K. Panda HINT: Designing Cache-Efficient MPI_Alltoall using Hybrid Memory Copy Ordering and Non-Temporal Instructions. 802-813 2024 IPDPS https://doi.org/10.1109/IPDPS57955.2024.00076 conf/ipps/2024 db/conf/ipps/ipdps2024.html#0005CASASSP24 Qinghua Zhou Bharath Ramesh 0005 Aamir Shafi Mustafa Abduljabbar Hari Subramoni Dhabaleswar K. Panda 0001 Accelerating MPI AllReduce Communication with Efficient GPU-Based Compression Schemes on Modern GPU Clusters. 1-12 2024 ISC https://doi.org/10.23919/ISC.2024.10528931 conf/supercomputer/2024 db/conf/supercomputer/isc2024.html#Zhou0SAS024

Kawthar Shafie Khorassani Chen-Chun Chen Bharath Ramesh 0005 Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 High Performance MPI over the Slingshot Interconnect. 128-145 2023 February 38 J. Comput. Sci. Technol. 1 https://doi.org/10.1007/s11390-023-2907-5 db/journals/jcst/jcst38.html#KhorassaniCRSSP23

Kaushik Kandadi Suresh Kawthar Shafie Khorassani Chen-Chun Chen Bharath Ramesh 0005 Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 Network-Assisted Noncontiguous Transfers for GPU-Aware MPI Libraries. 131-139 2023 March - April 43 IEEE Micro 2 https://doi.org/10.1109/MM.2023.3241133 db/journals/micro/micro43.html#SureshKCRASSP23

Bharath Ramesh 0005 Goutham Kalikrishna Reddy Kuncham Kaushik Kandadi Suresh Rahul Vaidya Nawras Alnaasan Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. D. K. Panda Designing In-network Computing Aware Reduction Collectives in MPI. 25-32 2023 HOTI https://doi.org/10.1109/HOTI59126.2023.00018 conf/hoti/2023 db/conf/hoti/hoti2023.html#RameshKSVAASSP23 Nicholas Contini Bharath Ramesh 0005 Kaushik Kandadi Suresh Tu Tran Benjamin Michalowicz Mustafa Abduljabbar Hari Subramoni Dhabaleswar K. Panda 0001 Enabling Reconfigurable HPC through MPI-based Inter-FPGA Communication. 477-487 2023 ICS https://doi.org/10.1145/3577193.3593720 https://www.wikidata.org/entity/Q130808910 conf/ics/2023 db/conf/ics/ics2023.html#Contini0STMAS023 Kaushik Kandadi Suresh Benjamin Michalowicz Bharath Ramesh 0005 Nicholas Contini Jinghan Yao Shulei Xu Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 A Novel Framework for Efficient Offloading of Communication Operations to Bluefield SmartNICs. 123-133 2023 IPDPS https://doi.org/10.1109/IPDPS54959.2023.00022 conf/ipps/2023 db/conf/ipps/ipdps2023.html#SureshMRCYXSSP23 Benjamin Michalowicz Kaushik Kandadi Suresh Bharath Ramesh 0005 Aamir Shafi Hari Subramoni Mustafa Abduljabbar Dhabaleswar K. Panda 0001 In-Depth Evaluation of a Lower-Level Direct-Verbs API on InfiniBand-based Clusters: Early Experiences. 354-363 2023 IPDPS Workshops https://doi.org/10.1109/IPDPSW59300.2023.00065 conf/ipps/2023w db/conf/ipps/ipdps2023w.html#MichalowiczSRSSAP23 Bharath Ramesh 0005 Qinghua Zhou Aamir Shafi Mustafa Abduljabbar Hari Subramoni Dhabaleswar K. Panda 0001 Designing Efficient Pipelined Communication Schemes using Compression in MPI Libraries. 95-99 2022 HIPC https://doi.org/10.1109/HiPC56025.2022.00024 conf/hipc/2022 db/conf/hipc/hipc2022.html#RameshZSASP22 Kaushik Kandadi Suresh Akshay Paniraja Guptha Benjamin Michalowicz Bharath Ramesh 0005 Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 Efficient Personalized and Non-Personalized Alltoall Communication for Modern Multi-HCA GPU-Based Clusters. 100-104 2022 HIPC https://doi.org/10.1109/HiPC56025.2022.00025 conf/hipc/2022 db/conf/hipc/hipc2022.html#SureshGMRASSP22 Kaushik Kandadi Suresh Kawthar Shafie Khorassani Chen-Chun Chen Bharath Ramesh 0005 Mustafa Abduljabbar Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 Network Assisted Non-Contiguous Transfers for GPU-Aware MPI Libraries. 13-20 2022 HOTI https://doi.org/10.1109/HOTI55740.2022.00018 conf/hoti/2022 db/conf/hoti/hoti2022.html#SureshKCRASSP22 Tu Tran Benjamin Michalowicz Bharath Ramesh 0005 Hari Subramoni Aamir Shafi Dhabaleswar K. Panda 0001 Designing Hierarchical Multi-HCA Aware Allgather in MPI. 28:1-28:10 2022 ICPP Workshops https://doi.org/10.1145/3547276.3548524 https://www.wikidata.org/entity/Q130982572 conf/icppw/2022 db/conf/icppw/icppw2022.html#TranM0SS022 Kawthar Shafie Khorassani Chen-Chun Chen Bharath Ramesh 0005 Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 High Performance MPI over the Slingshot Interconnect: Early Experiences. 15:1-15:7 2022 PEARC https://doi.org/10.1145/3491418.3530773 https://www.wikidata.org/entity/Q130972125 conf/xsede/2022 db/conf/xsede/pearc2022.html#KhorassaniC0SS022 Bharath Ramesh 0005 Jahanzeb Maqbool Hashmi Shulei Xu Aamir Shafi Seyedeh Mahdieh Ghazimirsaeed Mohammadreza Bayatpour Hari Subramoni Dhabaleswar K. Panda 0001 Towards Architecture-aware Hierarchical Communication Trees on Modern HPC Systems. 272-281 2021 HiPC https://doi.org/10.1109/HiPC53243.2021.00041 conf/hipc/2021 db/conf/hipc/hipc2021.html#0005HXSGBS021 Kaushik Kandadi Suresh Bharath Ramesh 0005 Chen-Chun Chen Seyedeh Mahdieh Ghazimirsaeed Mohammadreza Bayatpour Aamir Shafi Hari Subramoni Dhabaleswar K. Panda 0001 Layout-aware Hardware-assisted Designs for Derived Data Types in MPI. 302-311 2021 HiPC https://doi.org/10.1109/HiPC53243.2021.00044 conf/hipc/2021 db/conf/hipc/hipc2021.html#Suresh0CGBSS021 Nick Sarkauskas Mohammadreza Bayatpour Tu Tran Bharath Ramesh 0005 Hari Subramoni Dhabaleswar K. Panda 0001 Large-Message Nonblocking MPI_Iallgather and MPI Ibcast Offload via BlueField-2 DPU. 388-393 2021 HiPC https://doi.org/10.1109/HiPC53243.2021.00054 conf/hipc/2021 db/conf/hipc/hipc2021.html#SarkauskasBT0S021 Jahanzeb Maqbool Hashmi Shulei Xu Bharath Ramesh 0005 Mohammadreza Bayatpour Hari Subramoni Dhabaleswar K. D. K. Panda Machine-agnostic and Communication-aware Designs for MPI on Emerging Architectures. 32-41 2020 IPDPS https://doi.org/10.1109/IPDPS47924.2020.00014 conf/ipps/2020 db/conf/ipps/ipdps2020.html#HashmiX0BSP20 Kaushik Kandadi Suresh Bharath Ramesh 0005 Seyedeh Mahdieh Ghazimirsaeed Mohammadreza Bayatpour Jahanzeb Maqbool Hashmi Hari Subramoni Dhabaleswar K. D. K. Panda Performance Characterization of Network Mechanisms for Non-Contiguous Data Transfers in MPI. 896-905 2020 IPDPS Workshops https://doi.org/10.1109/IPDPSW50202.2020.00150 conf/ipps/2020w db/conf/ipps/ipdps2020w.html#Suresh0GBHSP20 Bharath Ramesh 0005 Kaushik Kandadi Suresh Nick Sarkauskas Mohammadreza Bayatpour Jahanzeb Maqbool Hashmi Hari Subramoni Dhabaleswar K. Panda 0001 Scalable MPI Collectives using SHARP: Large Scale Performance Evaluation on the TACC Frontera System. 11-20 2020 ExaMPI@SC https://doi.org/10.1109/ExaMPI52011.2020.00007 conf/sc/2020exampi db/conf/sc/exampi2020.html#0005SSBHSP20 Mohammadreza Bayatpour Jahanzeb Maqbool Hashmi Sourav Chakraborty 0003 Kaushik Kandadi Suresh Seyedeh Mahdieh Ghazimirsaeed Bharath Ramesh 0005 Hari Subramoni Dhabaleswar K. Panda 0001 Communication-Aware Hardware-Assisted MPI Overlap Engine. 517-535 2020 ISC https://doi.org/10.1007/978-3-030-50743-5_26 conf/supercomputer/2020 db/conf/supercomputer/isc2020.html#BayatpourH0SG0S20 Pouya Kousha Bharath Ramesh 0005 Kaushik Kandadi Suresh Ching-Hsiang Chu Arpan Jain Nick Sarkauskas Hari Subramoni Dhabaleswar K. Panda 0001 Designing a Profiling and Visualization Tool for Scalable and In-depth Analysis of High-Performance GPU Clusters. 93-102 2019 HiPC https://doi.org/10.1109/HiPC.2019.00022 conf/hipc/2019 db/conf/hipc/hipc2019.html#KoushaRSCJSSP19 Amit Ruhela Bharath Ramesh 0005 Sourav Chakraborty 0003 Hari Subramoni Jahanzeb Maqbool Hashmi Dhabaleswar K. Panda 0001 Leveraging Network-level parallelism with Multiple Process-Endpoints for MPI Broadcast. 34-41 2019 IPDRM@SC https://doi.org/10.1109/IPDRM49579.2019.00009 conf/sc/2019ipdrm db/conf/sc/ipdrm2019.html#Ruhela00SHP19 Hooyoung Ahn Shin-Young AhnShinyoung Ahn Nawras Alnaasan Mohammadreza Bayatpour Sourav Chakraborty 0003 Chen-Chun Chen Ching-Hsiang Chu Nicholas Contini Nick Contini Seyedeh Mahdieh Ghazimirsaeed Akshay Paniraja Guptha Woojong Han Jahanzeb Maqbool Hashmi Mustafa Abdul JabbarMustafa Abduljabbar Arpan Jain Kawthar Shafie Khorassani Seonyoung Kim Pouya Kousha Goutham Kalikrishna Reddy Kuncham Benjamin Michalowicz Dhabaleswar K. Panda 0001Dhabaleswar K. D. K. Panda Yoo-Mi Park Jake Queiser Amit Ruhela Nick Sarkauskas Aamir Shafi Hari Subramoni Kaushik Kandadi Suresh Tu Tran Rahul Vaidya Shulei Xu Jinghan Yao Qinghua Zhou