John Ravi

College of Engineering

Works (9)

Updated: September 19th, 2023 05:01

2023 article

Evaluating Asynchronous Parallel I/O on HPC Systems

2023 IEEE INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM, IPDPS, pp. 211–221.

By: J. Ravi n, S. Byna *, Q. Koziol *, H. Tang * & M. Becchi n

author keywords: Performance Evaluation; Modeling; Asynchronous; I/O; Parallel I/O
TL;DR: A systematic study of various factors affecting the performance and efficacy of asynchronous I/O in HPC systems is performed, a performance model is developed to estimate the aggregateI/O bandwidth achievable by iterative applications using synchronous and asynchronous I-O based on past observations, and the performance of the recently developed asynchronous I /O feature of a parallel I/o library (HDF5) is evaluated using benchmarks and real-world science applications. (via Semantic Scholar)
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: September 5, 2023

2023 article

Runway: In-transit Data Compression on Heterogeneous HPC Systems

2023 IEEE/ACM 23RD INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND INTERNET COMPUTING, CCGRID, pp. 229–239.

By: J. Ravi n, S. Byna * & M. Becchi n

author keywords: Object Data Management; In-transit Computation; Heterogeneous Resources
TL;DR: This paper introduces Runway, a runtime framework that enables computation on in-transit data with an object storage abstraction that is designed to be extensible to execute user-defined functions at runtime. (via Semantic Scholar)
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: August 21, 2023

2023 article

Runway: In-transit Data Compression on Heterogeneous HPC Systems

2023 IEEE/ACM 23RD INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND INTERNET COMPUTING WORKSHOPS, CCGRIDW, pp. 340–342.

By: J. Ravi n, S. Byna * & M. Becchi n

author keywords: Object Data Management; In-transit Computation; Heterogeneous Resources
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: September 18, 2023

2022 article

HDF5 Cache VOL: Efficient and Scalable Parallel I/O through Caching Data on Node-local Storage

2022 22ND IEEE/ACM INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND INTERNET COMPUTING (CCGRID 2022), pp. 61–70.

By: H. Zheng *, V. Vishwanath *, Q. Koziol *, H. Tang *, J. Ravi n, J. Mainzer*, S. Byna *

author keywords: Parallel IO; storage hierarchy; node-local storage; HDF5; Caching; Prefetching; deep learning
TL;DR: This paper designed this to move data asynchronously between the caching storage layer and a parallel file system to overlap the data movement overhead in performing I/O with compute phases, thus achieving faster time-to-solution in scientific simulations. (via Semantic Scholar)
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: September 29, 2022

2021 journal article

Molecular organization of cytokinesis node predicts the constriction rate of the contractile ring

JOURNAL OF CELL BIOLOGY, 220(3).

By: K. Bellingham-Johnstun n, E. Anders n, J. Ravi n, C. Bruinsma n & C. Laplante n

MeSH headings : Acetyltransferases / metabolism; Actin Cytoskeleton / metabolism; Actins / metabolism; Cytokinesis; Models, Biological; Mutation / genetics; Schizosaccharomyces / cytology; Schizosaccharomyces / metabolism; Schizosaccharomyces pombe Proteins / genetics; Schizosaccharomyces pombe Proteins / metabolism; Time Factors
TL;DR: This work leveraged the Δmid1 contractile ring assembly mechanism to determine how two distinct molecular organizations, nodes versus strands, can yield functional contractile rings and establishes a predictive correlation between the molecular organization of nodes and the behavior of the contractiles ring. (via Semantic Scholar)
Source: Web Of Science
Added: March 29, 2021

2021 article

PILOT: a Runtime System to Manage Multi-tenant GPU Unified Memory Footprint

2021 IEEE 28TH INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING, DATA, AND ANALYTICS (HIPC 2021), pp. 442–447.

By: J. Ravi n, T. Nguyen n , H. Zhou n  & M. Becchi n

TL;DR: A proposed three methods to transparently mitigate memory interference through kernel preemption and scheduling policies are proposed, which would enable new OS-managed scheduling policies to be implemented for GPU kernels to dynamically handle resource contention and offer consistent performance. (via Semantic Scholar)
Sources: Web Of Science, NC State University Libraries, ORCID
Added: May 2, 2022

2021 journal article

Transparent Asynchronous Parallel I/O Using Background Threads

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 33(4), 891–902.

By: H. Tang *, Q. Koziol *, J. Ravi n & S. Byna *

author keywords: Task analysis; Connectors; Libraries; Instruction sets; Computational modeling; Monitoring; Middleware; Asynchronous I; O; parallel I; O; background threads
TL;DR: An asynchronous I/O framework is presented that supports all types ofI/O operations, manages data dependencies transparently and automatically, provides implicit and explicit modes for application flexibility, and error information retrieval, and is implemented in HDF5. (via Semantic Scholar)
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: November 1, 2021

2020 article

GPU Direct I/O with HDF5

PROCEEDINGS OF 2020 IEEE/ACM FIFTH INTERNATIONAL PARALLEL DATA SYSTEMS WORKSHOP (PDSW 2020), pp. 28–33.

By: J. Ravi n, S. Byna * & Q. Koziol *

author keywords: GPU I/O; NVIDIA GPUDirect Storage (GDS); HDF5 GDS Virtual File Driver (VFD)
TL;DR: The effort of integrating GDS with HDF5, the top I/O library at NERSC and at DOE leadership computing facilities, is described and this integration using aHDF5 Virtual File Driver (VFD), which provides a file system abstraction to the application that allows HDF 5 applications to perform I/o without the need to move data between CPUs and GPUs explicitly. (via Semantic Scholar)
UN Sustainable Development Goal Categories
Source: Web Of Science
Added: August 9, 2021

2018 article

Compiling SIMT Programs on Multi- and Many-core Processors with Wide Vector Units: A Case Study with CUDA

2018 IEEE 25TH INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING (HIPC), pp. 123–132.

By: H. Wu n, J. Ravi n & M. Becchi n

author keywords: Xeon Phi; hybrid MIMD/SIMD systems; CUDA; SIMT; vectorization
TL;DR: A set of compiler techniques are proposed to transform programs written using a SIMT programming model (a subset of CUDA C) into code that leverages both the x86 cores and the vector units of a hybrid MIMD/SIMD architecture, thus providing programmability, high system utilization and performance. (via Semantic Scholar)
Source: Web Of Science
Added: June 17, 2019

Citation Index includes data from a number of different sources. If you have questions about the sources of data in the Citation Index or need a set of data which is free to re-distribute, please contact us.

Certain data included herein are derived from the Web of Science© and InCites© (2024) of Clarivate Analytics. All rights reserved. You may not copy or re-distribute this material in whole or in part without the prior written consent of Clarivate Analytics.