Skip to main content

Gnosis Research Center

From high-performance storage to autonomous science.

GRC develops memory, storage and I/O systems for high-performance computing. We work with national laboratories to power AI and agentic science.

Explore the research

Research areas

Gnosis Research Center develops computing systems for data-intensive science. Its research spans memory systems architecture, parallel and distributed I/O, resource management, and agentic science and AI.

  • Memory Systems Architecture

    Memory hierarchy modeling, cache management, and compile- and run-time tools for high-performance systems.

  • Parallel and Distributed I/O

    I/O buffering, parallel file systems, and data movement across deep memory and storage hierarchies.

  • Resource Management

    Task scheduling, data placement, and fast data access for high-end computing.

  • Agentic Science and AI

    Context engineering and agent coordination for autonomous discovery: the CLIO architecture within IOWarp, memory layers for multi-agent systems, and AI-driven scientific workflows.

Featured projects

All projects

FlagshipActive since 2024

IOWarp

Context management for AI-assisted scientific workflows across heterogeneous storage

30 publications · 18 contributors · NSF OAC-2411318 (opens in a new tab)

Explore IOWarp

ChronoLog

A time-ordered shared log with concurrent access and tiered capacity

Active since 2021

LABIOS

A patented label-based I/O paradigm, rebuilt as an agent I/O runtime for HPC applications and AI agents

Active since 2017

DTIO

A task-based I/O runtime unifying HPC, Big Data, and ML data stacks through the DataTask abstraction

Active since 2023

Latest from the center

Software releases

All projects

Go deeper

Teaching

Tutorials & Courses

The center teaches what it builds: hands-on tutorials and course material on HPC storage, parallel I/O, and the systems behind its research.

Start learning

Infrastructure

The Ares Cluster

A 660-core research cluster with node-local NVMe, SSD, and HDD storage, built for I/O-intensive experiments and open to the center's students and collaborators.

See the hardware

Students

PAIDEIA

The center's AI-fluency training framework: a structured pathway for researchers from first exposure to agent orchestration.

Read the thesis

Intellectual property

Patents & IP

Issued patents from GRC and its predecessor, the SCS Lab, covering hierarchical buffering, data placement, and memory systems.

Browse the portfolio

Join the center

A center that builds what it publishes.

GRC trains graduate researchers and works alongside national laboratories, federal agencies, and industry on the systems that carry science’s data. See the current openings and contact the center about research partnerships.

View open positionsPartner with GRC

Prospective students: start with For students in Resources for research pathways, mentorship, and how to join the lab.

The GRC team at the center's research booth at SC23
The center's team at SC23, Denver

Sponsors & partners

Partner network →
  • NVIDIA
  • The HDF Group
  • Department of Energy
  • National Aeronautics and Space Administration
  • National Institutes of Health
  • National Science Foundation
  • Argonne Leadership Computing Facility (ALCF)
  • Argonne National Laboratory
  • Chameleon Cloud Testbed
  • Lawrence Berkeley National Laboratory
  • Lawrence Livermore National Laboratory
  • Los Alamos National Laboratory
  • National Center for Supercomputing Applications (NCSA)
  • National Energy Research Scientific Computing Center
  • Oak Ridge Leadership Computing Facility (OLCF)
  • Oak Ridge National Laboratory
  • Pacific Northwest National Laboratory
  • Sandia National Laboratories
  • DePaul University
  • Northwestern University
  • Texas Tech University
  • University of California, Santa Cruz
  • University of Chicago
  • University of Illinois Chicago
  • University of Illinois Urbana-Champaign
  • University of Piraeus, Greece
  • University of Tennessee, Knoxville
  • University of Utah
  • Google Cloud Platform
  • MLCommons