Hariharan Devarajan
Portrait of Hariharan Devarajan

Hariharan Devarajan

Computer Scientist · Center for Applied Scientific Computing (CASC), Lawrence Livermore National Laboratory

Explainable I/O, data movement, and storage systems for scientific and AI workflows

I am a Computer Scientist at the Center for Applied Scientific Computing (CASC), Lawrence Livermore National Laboratory (LLNL). I work on data movement, storage systems, and performance analysis for large-scale scientific and AI workflows on supercomputers. My research makes I/O explainable: connecting low-level storage telemetry with workflow context, tracking the lineage of scientific data, and turning those insights into targeted optimizations. I lead the Moirai provenance project (LDRD, PI), co-lead DLDL (deep learning data loading), lead data management for the Heuretes strategic initiative, and created the DLIO Benchmark (part of MLPerf Storage) and DFTracer. I received my Ph.D. in Computer Science from Illinois Tech.

CV (PDF) EmailLLNL ProfileGitHubLinkedInGoogle ScholarORCID
49publications
29conference papers
2best paper awards
R&D 100award for UnifyFS

Research

Making data movement in HPC and AI workflows observable, explainable, and fast.

Explainable I/O

Explaining why I/O is slow by connecting storage telemetry with workflow context and the anatomy of the storage stack: DataCrumbs and ORION (ProTools'26), Metis (REX-IO'26), and XIO (SSDBM'25).

Performance Monitoring and Analysis

Scalable tracing and automated analysis of workflow data movement: DFTracer (SC'24), DFAnalyzer, WisIO (ICS'25), HORATIO (SSDBM'26), IOMax, Recorder, and SmartIO.

Provenance for Scientific Workflows

Moirai, an LDRD project I lead as PI, tracks the lineage and lifecycle of data across workflow tasks, software layers, and storage systems.

Deep Learning I/O

From monitoring to characterization to optimization of AI training pipelines: DLIO (CCGrid'21 Best Paper; MLPerf Storage), GLANCED-IO (HPDC'26), CoDL (REX-IO'26), Stimulus, and DYAD (SBAC-PAD'24).

Workload-aware Middleware

Hermes (HPDC'18), UnifyFS (IPDPS'26; R&D 100), DYAD, WADO (SSDBM'26), HCL, LABIOS (HPDC'19 Best Paper), and intent-driven I/O (Mimir, H5Intent, IFlux, iCR).

Projects

Current and recent projects and my role in each.

Principal Investigator

Moirai

Trusted data provenance for HPC workflows (LDRD Exploratory Research). Tracks the lineage and lifecycle of data across scientific workflows.

Co-Principal Investigator

DLDL

Deep learning data loading: benchmarking, monitoring, and optimizing deep learning I/O pipelines on HPC systems with DLIO and DFTracer.

Data Management Thrust Lead

Heuretes

Cross-site data movement and scheduling under time constraints (LDRD Strategic Initiative).

Lead

XIO

I/O explainability and the anatomy of HPC storage stacks with DataCrumbs, ORION, and Metis.

Lead Developer and Maintainer

GOTCHA

Mission-critical LLNL function-interposition library supporting tools across the DOE and NNSA ecosystem.

Data Management Thrust Lead (past)

FRACTALE

Multi-cluster scheduling and data management; built IFlux and DataCrumbs.

Code

GitHubLLNL ASRBitbucket

Publications

49 papers, newest first. Filter by type or search.

2026
  1. DataCrumbs: Efficient Cross-Layer Data Capture for Explainable I/O Performance in HPC Storage Stacks
    Hariharan Devarajan, Sarah Neuwirth, Chen Wang, Ray A. O. Sinurat, Neeraj Rajesh, Jay Lofstead, Vanessa Sochat, Daniel Milroy, and Thomas Scogland
    Proceedings of the SC'26 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC Workshops'26): 8th Workshop on Programming and Performance Visualization Tools (ProTools'26), Chicago, IL, USA, IEEE, Nov. 2026
    WorkshopPDF
  2. ORION: Observability Framework for Hybrid Workflow-Centric I/O Monitoring in HPC Systems
    Hariharan Devarajan, Ray A. O. Sinurat, Izzet Yildirim, Olga Kogiou, Aniruddha Marathe, and Nikoli Dryden
    Proceedings of the SC'26 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC Workshops'26): 8th Workshop on Programming and Performance Visualization Tools (ProTools'26), Chicago, IL, USA, IEEE, Nov. 2026
    WorkshopPDF
  3. iCR: Intent-Driven Adaptable Checkpoint/Restart
    Olga Kogiou, Hariharan Devarajan, Chen Wang, Weikuan Yu, and Kathryn Mohror
    2026 IEEE International Conference on Cluster Computing (CLUSTER'26), Alexandria, VA, USA, IEEE, Sep. 2026
    ConferencePDFSlides
  4. WADO: A Distributed WORM Storage Service for Asynchronous Data Operations
    Karim Youssef, Hariharan Devarajan, Nikoli Dryden, and Roger Pearce
    Proceedings of the 38th International Conference on Scalable Scientific Data Management (SSDBM'26), ACM, Aug. 2026, pp. 1–12
    ConferencePDF
  5. HORATIO: Bridging Management and Analysis of Traces at Scale
    Ray A. O. Sinurat, Hariharan Devarajan, Nikoli Dryden, Haryadi S. Gunawi, and William Nixon
    Proceedings of the 38th International Conference on Scalable Scientific Data Management (SSDBM'26), ACM, Aug. 2026, pp. 1–12
    ConferencePDFSlides
  6. GLANCED-IO: Taming I/O Optimization for Deep Learning at Scale
    Ray A. O. Sinurat, William Nixon, Philip Carns, Huihuo Zheng, Sandeep Madireddy, Sam Foreman, Troy Arcomano, Robert B. Ross, Haryadi S. Gunawi, and Hariharan Devarajan
    Proceedings of the 35th International Symposium on High-Performance Parallel and Distributed Computing (HPDC'26), ACM, Jul. 2026, pp. 305–318
    ConferencePDFSlides
  7. Metis: Agentic Knowledge Synthesis for Explainable I/O Performance in HPC Systems
    Karim Youssef, Sarah Neuwirth, Neeraj Rajesh, and Hariharan Devarajan
    6th Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO'26), ACM, Jul. 2026
    WorkshopPDF
  8. CoDL: A Framework for Studying Cross-Component Interference in Deep Learning Training Pipelines
    Druva Dhakshinamoorthy, Ray A. O. Sinurat, Nikoli Dryden, Arnab K. Paul, and Hariharan Devarajan
    6th Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO'26), ACM, Jul. 2026
    WorkshopPDFSlides
  9. Chasing the Rabbit: A Systematic Exploration of Software-Defined HPC Storage
    Hariharan Devarajan, Brian Behlendorf, Blake Devcich, Dean Roehrich, Matt Richerson, Tony Floeder, Brent Petit, Don Boyd, Trace Hamada, Tim McCree, Gian-Carlo Defazio, James Corbett, Elsa Gonsiorowski, Cameron Stanavige, Nikoli Dryden, Roger Pearce, Kathryn Mohror, and Bronis R. de Supinski
    Proceedings of the 7th Workshop on Extreme-Scale Storage and Analysis (ESSA'26), held in conjunction with the 40th IEEE International Parallel and Distributed Processing Symposium (IPDPS'26), IEEE, May 2026
    WorkshopPDFSlides
  10. IFlux: Intent-Aware Storage Tiers & Software Scheduling for HPC Systems
    Hariharan Devarajan, Vanessa Sochat, Daniel J. Milroy, and Tom Scogland
    IEEE Transactions on Parallel and Distributed Systems, 2026, pp. 2236–2251
    JournalPDF
2025
  1. SmartIO: A Lightweight End-to-End Workflow for Runtime I/O Optimization of HPC Systems
    Hammad Bin Ather, Chen Wang, Hariharan Devarajan, Hank Childs, and Kathryn Mohror
    Proceedings of the Parallel Data Storage Workshop (PDSW'25), held in conjunction with SC25: The International Conference for High Performance Computing, Networking, Storage and Analysis, ACM, Nov. 2025, pp. 1396–1405
    WorkshopPDF
  2. Optimizing I/O for an Exascale Implicit Kinetic Plasma Simulation using the Rabbit Storage System
    Ian Lumsden, Hariharan Devarajan, Izzet Yildirim, Stefano Markidis, Andong Hu, Ivy Bo Peng, Luca Pennati, Dewi Yokelson, Stephanie Brink, Olga Pearce, Tom Scogland, Bronis R. de Supinski, Gian Luca Delzanno, Anthony Kougkas, Xian-He Sun, and Michela Taufer
    5th Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO'25), 2025 IEEE International Conference on Cluster Computing Workshops (CLUSTER Workshops), IEEE, Sep. 2025, pp. 1–6
    WorkshopLLNL Director's Institutional AwardPDF
  3. Performance Optimization of an Exascale Implicit Kinetic Plasma Simulation on El Capitan
    Ian Lumsden, Stefano Markidis, Andong Hu, Ivy Bo Peng, Luca Pennati, Dewi Yokelson, Stephanie Brink, Olga Pearce, Tom Scogland, Hariharan Devarajan, Bronis R. de Supinski, Gian Luca Delzanno, and Michela Taufer
    2025 IEEE International Conference on eScience (eScience), IEEE, Sep. 2025, pp. 319–320
    Poster
  4. WisIO: Automated I/O Bottleneck Detection with Multi-Perspective Views for HPC Workflows
    Izzet Yildirim, Hariharan Devarajan, Anthony Kougkas, Xian-He Sun, and Kathryn Mohror
    Proceedings of the 39th ACM International Conference on Supercomputing (ICS'25), ACM, Jun. 2025, pp. 749–763
    ConferencePDF
  5. XIO: Toward eXplainable I/O for HPC Systems
    Sarah Neuwirth, Hariharan Devarajan, Chen Wang, and Jay Lofstead
    Proceedings of the 37th International Conference on Scalable Scientific Data Management (SSDBM'25), ACM, Jun. 2025, pp. 1–6
    ConferencePDF
  6. Recorder: Comprehensive Parallel I/O Tracing and Analysis
    Chen Wang, Izzet Yildirim, Hariharan Devarajan, Kathryn Mohror, and Marc Snir
    arXiv:2501.04654, Jan. 2025
    PreprintPDF
2024
  1. ML-based Modeling to Predict I/O Performance on Different Storage Sub-systems
    Yiheng Xu, Pranav Sivaraman, Hariharan Devarajan, Kathryn Mohror, and Abhinav Bhatele
    31st IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), IEEE, Dec. 2024
    ConferencePDF
  2. DFTracer: An Analysis-Friendly Data Flow Tracer for AI-Driven Workflows
    Hariharan Devarajan, Loic Pottier, Kaushik Velusamy, Huihuo Zheng, Izzet Yildirim, Olga Kogiou, Weikuan Yu, Anthony Kougkas, Xian-He Sun, Jae Seung Yeom, and Kathryn Mohror
    SC24: International Conference for High Performance Computing, Networking, Storage and Analysis, Atlanta, GA, IEEE, Nov. 2024
    ConferencePDF
  3. TailorFS: An Adaptive file system to support dynamic I/O requirements of HPC workloads
    Hariharan Devarajan and Kathryn Mohror
    2024 IEEE 36th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), IEEE, Nov. 2024, pp. 81–92
    ConferencePDF
  4. DYAD: Locality-aware Data Management for accelerating Deep Learning Training
    Hariharan Devarajan, Ian Lumsden, Chen Wang, Konstantia Georgouli, Tom Scogland, Jae-Seung Yeom, and Michela Taufer
    2024 IEEE 36th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), IEEE, Nov. 2024, pp. 13–24
    ConferencePDF
  5. Understanding Highly Configurable Storage for Diverse Workloads
    Olga Kogiou, Hariharan Devarajan, Chen Wang, Weikuan Yu, and Kathryn Mohror
    4th Workshop on Re-envisioning Extreme-Scale I/O for Emerging Hybrid HPC Workloads (REX-IO), IEEE, Sep. 2024
    WorkshopPDF
  6. The impact of asynchronous I/O in checkpoint-restart workloads
    Hariharan Devarajan, Adam Moody, Donglai Dai, Cameron Stanavige, Elsa Gonsiorowski, Marty McFadden, Olaf Faaland, Greg Kosinovsky, and Kathryn Mohror
    5th Workshop on Extreme-Scale Storage and Analysis (ESSA), IEEE, May 2024
    WorkshopPDF
  7. Empirical Study of Molecular Dynamics Workflow Data Movement: DYAD vs. Traditional I/O Systems
    Ian Lumsden, Hariharan Devarajan, Jack Marquez, Stephanie Brink, David Boehme, Olga Pearce, Jae-Seung Yeom, and Michela Taufer
    23rd IEEE International Workshop on High Performance Computational Biology (HICOMB), IEEE, May 2024
    WorkshopPDF
  8. H5Intent: Autotuning HDF5 with User Intent
    Hariharan Devarajan, Gerd Heber, and Kathryn Mohror
    IEEE Transactions on Parallel and Distributed Systems, 2024
    JournalPDF
2023
  1. IOMax: Maximizing Out-of-Core I/O Analysis Performance on HPC Systems
    Izzet Yildirim, Hariharan Devarajan, Anthony Kougkas, Xian-He Sun, and Kathryn Mohror
    Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis (SC-W 2023), Denver, CO, USA, ACM, Nov. 2023, pp. 1209–1215
    WorkshopPDF
  2. I/O characterization and performance evaluation of large-scale storage architectures for heterogeneous workloads
    Olga Kogiou, Hariharan Devarajan, Chen Wang, Weikuan Yu, and Kathryn Mohror
    2023 IEEE International Conference on Cluster Computing Workshops (CLUSTER Workshops), IEEE, Oct. 2023, pp. 44–45
    Poster
  3. Mimir: Extending I/O Interfaces to Express User Intent for Complex Workloads in HPC
    Hariharan Devarajan and Kathryn Mohror
    2023 IEEE International Parallel and Distributed Processing Symposium (IPDPS), St. Petersburg, Florida, IEEE, May 2023
    ConferencePDF
2022
  1. Extracting and characterizing I/O behavior of HPC workloads
    Hariharan Devarajan and Kathryn Mohror
    2022 IEEE International Conference on Cluster Computing (CLUSTER'22), Heidelberg, Germany, Sep. 2022
    ConferencePDF
  2. Stimulus: Accelerate Data Management for Scientific AI applications in HPC
    Hariharan Devarajan, Anthony Kougkas, Huihuo Zheng, Venkatram Vishwanath, and Xian-He Sun
    2022 IEEE/ACM International Symposium in Cluster, Cloud, and Internet Computing (CCGrid'22), Taormina, Italy, May 2022
    ConferencePDF
2021
  1. HFlow: A Dynamic and Elastic Multi-Layered Data Forwarder
    Jaime Cernuda Garcia, Hariharan Devarajan, Luke Logan, Keith Bateman, Neeraj Rajesh, Jie Ye, Anthony Kougkas, and Xian-He Sun
    2021 IEEE International Conference on Cluster Computing (CLUSTER'21), Portland, OR, IEEE, Sep. 2021
    ConferencePDF
  2. Apollo: An ML-assisted Real-Time Storage Resource Observer
    Neeraj Rajesh, Hariharan Devarajan, Jaime Cernuda Garcia, Keith Bateman, Luke Logan, Jie Ye, Anthony Kougkas, and Xian-He Sun
    Proceedings of the 30th International Symposium on High-Performance Parallel and Distributed Computing (HPDC'21), New York, NY, USA, Association for Computing Machinery, Jun. 2021, pp. 147–159
    ConferencePDF
  3. DLIO: A Data-Centric Benchmark for Scientific Deep Learning Applications
    Hariharan Devarajan, Huihuo Zheng, Anthony Kougkas, Xian-He Sun, and Venkatram Vishwanath
    2021 21st IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGRID), IEEE, May 2021
    ConferenceBest PaperPDF
2020
  1. HReplica: A Dynamic Data Replication Engine with Adaptive Compression for Multi-Tiered Storage
    Hariharan Devarajan, Anthony Kougkas, and Xian-He Sun
    2020 IEEE International Conference on Big Data (Big Data), Atlanta, Georgia, USA, Dec. 2020
    ConferencePDF
  2. HCL: Distributing Parallel Data Structures in Extreme Scales
    Hariharan Devarajan, Anthony Kougkas, Keith Bateman, and Xian-He Sun
    2020 IEEE International Conference on Cluster Computing (CLUSTER), IEEE, Sep. 2020
    ConferencePDFSlides
  3. HCompress: Hierarchical Data Compression for Multi-Tiered Storage Environments
    Hariharan Devarajan, Anthony Kougkas, Luke Logan, and Xian-He Sun
    2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS), New Orleans, Louisiana, USA, May 2020
    ConferencePDF
  4. HFetch: Hierarchical Data Prefetching for Scientific Workflows in Multi-Tiered Storage Environments
    Hariharan Devarajan, Anthony Kougkas, and Xian-He Sun
    2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS), New Orleans, Louisiana, USA, May 2020
    ConferencePDF
  5. Bridging Storage Semantics using Data Labels and Asynchronous I/O
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    ACM Transactions on Storage (TOS), 2020
    ConferencePDF
  6. ChronoLog: A Distributed Shared Tiered Log Store with Time-based Data Ordering
    Anthony Kougkas, Hariharan Devarajan, Keith Bateman, Jaime Cernuda, Neeraj Rajesh, and Xian-He Sun
    Proceedings of the 36th International Conference on Massive Storage Systems and Technology (MSST 2020), 2020
    ConferencePDF
2019
  1. NIOBE: An Intelligent I/O Bridging Engine for Complex and Distributed Workflows
    K. Feng, H. Devarajan, A. Kougkas, and X.-H. Sun
    IEEE International Conference on Big Data, Dec. 2019
    ConferencePDF
  2. LABIOS: A Distributed Label-Based I/O System
    Anthony Kougkas, Hariharan Devarajan, Jay Lofstead, and Xian-He Sun
    Proceedings of the ACM 28th International Symposium on High-Performance Parallel and Distributed Computing (HPDC'19), Jun. 2019
    ConferenceBest PaperPDFSlides
  3. An Intelligent, Adaptive, and Flexible Data Compression Framework
    Hariharan Devarajan, Anthony Kougkas, and Xian-He Sun
    Proceedings of the IEEE/ACM International Symposium in Cluster, Cloud, and Grid Computing (CCGrid'19), May 2019
    ConferencePDF
  4. I/O Acceleration via Multi-Tiered Data Buffering and Prefetching
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    Journal of Computer Science and Technology, 2019
    JournalPDF
2018
  1. Vidya: Performing Code-Block I/O Characterization for Data Access Optimization
    Hariharan Devarajan, Anthony Kougkas, Prajwal Challa, and Xian-He Sun
    Proceedings of the IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), Dec. 2018
    ConferencePDF
  2. Harmonia: An Interference-Aware Dynamic I/O Scheduler for Shared Non-Volatile Burst Buffers
    Anthony Kougkas, Hariharan Devarajan, Xian-He Sun, and Jay Lofstead
    Proceedings of the IEEE International Conference on Cluster Computing (Cluster'18), Sep. 2018
    ConferencePDFSlides
  3. Hermes: A Heterogeneous-Aware Multi-Tiered Distributed I/O Buffering System
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    Proceedings of the ACM 27th International Symposium on High-Performance Parallel and Distributed Computing (HPDC'18), Jun. 2018
    ConferencePDF
  4. IRIS: I/O Redirection via Integrated Storage
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    Proceedings of the 32nd ACM International Conference on Supercomputing (ICS'18), Jun. 2018
    ConferencePDF
2017
  1. Enosis: Bridging the Semantic Gap between File-based and Object-based Data Models
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    Proceedings of the ACM SIGHPC Datacloud'17, 8th International Workshop on Data-Intensive Computing in the Clouds, Nov. 2017
    WorkshopPDF
  2. Syndesis: Mapping Objects to Files for a Unified Data Access System
    Anthony Kougkas, Hariharan Devarajan, and Xian-He Sun
    Proceedings of the ACM SIGHPC MTAGS'17, 10th International Workshop on Many-Task Computing on Clouds, Grids, and Supercomputers, in conjunction with SC'17, Nov. 2017
    WorkshopPDF
  3. Open Ethernet Drive: Evolution of Energy-Efficient Storage Technology
    Hariharan Devarajan, Anthony Kougkas, Hsing-Bung Chen, and Xian-He Sun
    Proceedings of the ACM SIGHPC Datacloud'17, 8th International Workshop on Data-Intensive Computing in the Clouds, in conjunction with SC'17, Nov. 2017
    WorkshopPDF

No publications match.

Highlights

Awards, service, and recent talks.

Awards

  • Proposal award: "DLDL: Deep Learning Data Loading" (Co-Principal Investigator), 2025
  • LDRD Exploratory Research award: "Trusted Data Provenance for HPC Workflows (Moirai)" (Principal Investigator, 27-ERD-025), 2026
  • LDRD Strategic Initiative award: "FRACTALE: multi-cluster scheduling and data management" (Thrust Lead for Data Management), 2023
  • LDRD Strategic Initiative award: "Heuretes: cross-site data movement under time constraints" (Thrust Lead for Data Management), 2026
  • LLNL Director's Institutional Award for optimizing iPIC3D I/O on El Capitan with the Rabbit storage system (REX-IO'25)
  • R&D 100 Award for UnifyFS (2024)
  • Best Paper Award, IEEE/ACM CCGrid 2021 (DLIO)
  • Karsten Schwan Best Paper Award, ACM HPDC 2019 (LABIOS)
  • Outstanding Service Award, IEEE Cluster 2021

Service

  • Program Co-Chair, SSDBM 2026
  • Posters Co-Chair, HPDC 2026 and IEEE Cluster 2026
  • Program Committee: IPDPS, HPDC, Cluster, CCGrid, ICS, ICPP, ESSA, and REX-IO 2026
  • Review Board Member, IEEE Transactions on Parallel and Distributed Systems
  • Judge, IPDPS 2026 PhD Forum and PACT ACM Student Research Competition

Talks and Panels

  • Demo, "Chasing the Rabbit: Software-Defined Storage in Action," DOE booth at SC26
  • Talk, "Explaining HDF5 Performance through the I/O Stack," HDF5 BoF at SC26
  • Panelist, "Who Gets to Do Computing Research in 2036?" PACT 2026
  • Talk, "Skills Before Smarts: Harness Engineering for Robust Software-Development Agents," MLAC at PACT 2026 Slides
  • Keynote, "Why is my I/O Slow? Rethinking Storage Systems for Explainability," REX-IO'26 at HPDC 2026 Slides
  • Panelist, "Data Centers: Information, Misinformation, and the Future of Large-Scale Computing," HPDC 2026

Contact

Email

hariharandev1@llnl.gov

Office

Lawrence Livermore National Laboratory, 7000 East Ave, Livermore, CA 94550