Skip to main content
Research

Research Portfolio

Explore GRC systems for scientific AI, storage, memory, I/O, and distributed workflows, together with the results that shaped them.
IOWarp

IOWarp: Context Engineering Platform for Autonomous Scientific AI

Context-management infrastructure for organizing scientific data, metadata, telemetry, and computational state. CLIO Core provides storage and runtime services; CLIO Kit connects agents to scientific tools.

Funded by
Funded by NSFNational Science Foundation2024–2029
ChronoLog

ChronoLog: A Distributed Shared Log Store

A distributed shared log that orders activity and provenance data with physical time and moves records across memory, local storage, and archival tiers.

Funded by
Funded by NSFNational Science Foundation2021–2026
LABIOS

LABIOS: A Distributed Label-Based I/O System

LABIOS represents I/O as labels that carry an operation, data pointer, and routing metadata. Its public prototype routes labels from applications and agents through a dispatcher to Databot and Pipeline workers; the Agentic tier remains planned.

Funded by
Funded by NSFNational Science Foundation2023–2026

31 projects

15 current projects; 16 completed projects below.

IOWarp

Active

Context-management infrastructure for organizing scientific data, metadata, telemetry, and computational state. CLIO Core provides storage and runtime services; CLIO Kit connects agents to scientific tools.

2024–
Source code 30 publications 18 people

ChronoLog

Active

A distributed shared log that orders activity and provenance data with physical time and moves records across memory, local storage, and archival tiers.

2021–
Source code 19 publications 12 people

LABIOS

Active

LABIOS represents I/O as labels that carry an operation, data pointer, and routing metadata. Its public prototype routes labels from applications and agents through a dispatcher to Databot and Pipeline workers; the Agentic tier remains planned.

2017–
Source code 1 patent 13 publications 16 people

Hermes

Active

Distributed I/O buffering middleware that coordinates data placement across DRAM, NVMe, burst buffers, and parallel file systems.

2018–
Source code 30 publications 12 people

DTIO

Active

A task-based I/O runtime developed with Argonne National Laboratory to connect HPC, Big Data, and ML data stacks. The SSDBM'25 paper reports 49.6% better I/O performance for online translation with DataTask caching than for offline translation.

2023–
Source code 8 publications 9 people

Coeus

Active

An active-storage framework that computes derived quantities in transit and queries enriched metadata. The Hades paper reports 3 to 4 times faster analysis on tested Gray-Scott workflows.

2022–
Source code 9 publications 10 people

DeepIO

Active

Data-path methods for scientific AI, from the DLIO benchmark and Viper model transfer framework to UnboxKV characterization and PKAS scheduling for KV caches under concurrent LLM inference.

2020–
13 publications 12 people

Portus

Active

A DOE Genesis Mission project led by Pacific Northwest National Laboratory that ports scientific workflows to HPC systems and tunes where their tasks and data run. GRC contributes workflow performance research and CLIO.

2026–
6 publications 5 people

SPOTTER-AI

Active

A DOE Genesis Mission project led by Argonne National Laboratory that captures provenance across scientific workflows and traces and attributes threats to their data and results. GRC contributes CLIO from IOWarp.

2026–
4 people

WisIO

Active

WisIO analyzes HPC I/O traces through time, process, and file views, then connects metric-level bottlenecks with root causes.

2022–
Source code 9 publications 5 people

IRIS

Active

Unified data access middleware that translates between parallel file-system and object-store semantics.

2017–
Source code 20 publications 9 people

Memory Access Pattern Obfuscation

Active

This project combines cryptographic algorithms, application requirements, and memory architecture designs to protect memory access patterns from side channels.

2022–2027
7 publications 1 person

MPI4AI

Active

MPI4AI extends Open MPI with GPU communication, AI-oriented collectives, compute-stream integration, and fault tolerance for large-scale AI workloads.

2025–
1 publication 2 people

StoreHub

Winding Down

StoreHub combined two community instruments, a national workshop, and a landscape analysis to assess the need for dedicated data storage research infrastructure. Its public report recommends a staged federation built on existing facilities.

2024–
Source code 3 publications 2 people

UniMCC

Active

UniMCC coordinates architecture, code generation, runtime support, and performance models for near-memory processing and disaggregated memory.

2023–
7 publications 5 people
Archived / Completed Projects16
  • OptMemOptMem modeled data locality and memory-access concurrency together, then applied the model to adaptive prefetching and cache management.2020–2024
  • SMC2 PlanningSMC2 planning combines near-memory processors and a shared remote memory pool in a system design for graph-mining applications.2021–2022
  • IDESIDES integrates campus microgrid data with high-performance computing, storage, and networking for energy-system analytics and cybersecurity research.2017–2021
  • DiRecMRDiRecMR studies differences between map and reduce phases to improve task speculation, resource management, and resilience in MapReduce systems.2017–2018
  • Cloud LibraryCloud Library provided data management and visualization capabilities for cloud-resolving models running on Spark and Hadoop platforms, bridging scientific computing with big data frameworks.2014–2018
  • Integrated Data ManagementThis project bridged high-performance computing storage systems (Lustre, PVFS2, GPFS) with big data frameworks (HDFS, MapReduce), enabling unified data management across heterogeneous storage environments.2013–2017
  • DEPDEP proposed a decoupled execution paradigm that separates compute-intensive operations from data-intensive operations, enabling more efficient utilization of heterogeneous computing resources.2012–2016
  • Memory ParallelismThis project explored memory concurrency across hierarchy layers, developing techniques to exploit parallelism in modern memory systems for improved application performance.2010–2016
  • Push I/OPush I/O developed a server-push I/O architecture for application-specific optimization, conducted jointly with Argonne National Laboratory and the University of Illinois at Urbana-Champaign.2010–2015
  • Multicore SchedulingThis project developed core-aware scheduling algorithms, data access history caching, and prefetching mechanisms optimized for multicore processor architectures.2009–2015
  • FENCEFENCE was a hybrid fault-tolerance framework combining proactive fault avoidance with traditional checkpointing to improve resilience in high-performance computing environments.2007–2012
  • GHSGHS was a QoS-guaranteed task scheduling system for Grid computing, providing performance evaluation and resource harvesting capabilities. Source code was publicly released (v1.1).2002–2012
  • WorkflowThis project developed scientific workflow management capabilities for Lattice QCD computations on dedicated clusters, conducted jointly with Fermi National Accelerator Laboratory (Fermilab).2005–2010
  • HPCMHPCM was a heterogeneous process migration middleware enabling legacy code migration across different computing platforms, supporting mobility in high-performance computing environments.2000–2006
  • SNOWSNOW was a distributed metacomputing platform that enabled scalable computing across networks of heterogeneous workstations, laying the groundwork for later research in distributed systems and process migration.1999–2005
  • VCNSVCNS was a sister project of SNOW, providing a virtual collaboratory environment for numerical simulation across distributed computing resources.1999–2005

31 projects from 1999 to today across 12 themes. Choose a theme to filter the projects above.

Research themes by project, projects ordered by start year. A dot marks a theme the project carries; a filled dot is a live project, a ring a completed one.
1990s2000s2010s2020s
ThemeSNOWVCNSHPCMGHSWorkflowFENCEMulticore SchedulingMemory ParallelismPush I/ODEPIntegrated Data ManagementCloud LibraryDiRecMRIDESIRISLABIOSHermesDeepIOOptMemChronoLogSMC2 PlanningCoeusMemory Access Pattern ObfuscationWisIODTIOUniMCCIOWarpStoreHubMPI4AIPortusSPOTTER-AIProjects
yesyesyesyes4
yesyesyesyesyesyesyes7
yesyesyesyesyesyesyes7
yesyesyesyesyesyesyesyesyesyesyes11
yesyesyesyesyesyesyesyesyes9
yesyesyesyesyesyesyesyesyes9
yes1
yes1
yesyesyesyesyesyes6
yesyesyesyesyes5
yes1
yesyes2

Point at or tab to a project to read its record. Choose a theme to filter the projects above.

Live project Completed

  1. SNOWVCNSHPCMGHS

  2. WorkflowFENCEMulticore SchedulingDEPOptMemUniMCCMPI4AI

  3. Multicore SchedulingMemory ParallelismHermesOptMemSMC2 PlanningMemory Access Pattern ObfuscationUniMCC

  4. Push I/OIntegrated Data ManagementIRISLABIOSHermesDeepIOChronoLogCoeusDTIOIOWarpStoreHub

  5. Integrated Data ManagementCloud LibraryDiRecMRIDESIRISLABIOSChronoLogCoeusSPOTTER-AI

  6. IRISLABIOSHermesDeepIOChronoLogCoeusWisIODTIOPortus

  7. IDES

  8. DiRecMR

  9. DeepIOWisIODTIOIOWarpPortusSPOTTER-AI

  10. DeepIOIOWarpMPI4AIPortusSPOTTER-AI

  11. SMC2 Planning

  12. WisIOStoreHub