Haochen Pan

Thank you for stopping by! I am a fifth-year CS Ph.D. student at the University of Chicago and a proud member of Globus Labs (opens in new tab), advised by Dr. Kyle Chard (opens in new tab), Dr. Ian Foster (opens in new tab), and Dr. Ryan Chard (opens in new tab).
Email: haochenpan AT uchicago.edu
Haochen Pan

Current Research

My research focuses on distributed systems at the intersection of cloud and high-performance computing (HPC), with an emphasis on resilience and efficiency for AI-guided scientific workflows and time-sensitive data analysis.
  • We developed Octopus (opens in new tab), a Kafka-based hierarchical event fabric for high-performance exchange of control and metadata events across cloud and HPC environments.
  • Building on this, we designed Icicle (opens in new tab) (to appear at ISC High Performance 2026), a real-time metadata monitoring and indexing system for Lustre and IBM Storage Scale that integrates Octopus, Apache Flink, and Globus Search to provide live visibility and historical usage analysis.
  • Last year, we developed Science-MCP (opens in new tab), which applies the Model Context Protocol (MCP) to expose these capabilities as discoverable and composable services for LLM-powered agents across heterogeneous cyberinfrastructure.

Selected Publications

The complete list is available on Google Scholar (opens in new tab) and my CV.

Projects

All Publications

High-Performance Computing

  • Jun 2026[c23][ISC HPC'26 (opens in new tab)]Icicle: Scalable Metadata Indexing and Real-Time Monitoring for HPC File Systems
  • May 2026[j5][TMLCN (opens in new tab)]Throughput Estimation of Data Transport Networks from Digital Twin Measurements
  • Aug 2025[i3][Preprint (opens in new tab)]Experiences with Model Context Protocol Servers for Science and High Performance Computing
  • Aug 2025[j4][FHPCP (opens in new tab)]Toward a Persistent Event-Streaming System for High-Performance Computing Applications
  • Jul 2025[j3][ApJS (opens in new tab)]RADAR—Radio Afterglow Detection and AI-Driven Response: A Federated Framework for Gravitational Wave Event Follow-Up
  • Jun 2025[c22][ICS'25 (opens in new tab)]D-Rex: Heterogeneity-Aware Reliability Framework and Adaptive Algorithms for Distributed Storage
  • May 2025[c21][CCGrid'25 (opens in new tab)]DynoStore: A wide-area distribution system for the management of data over heterogeneous storage
  • May 2025[c20][CCGrid'25 (opens in new tab)]WRATH: Workload Resilience Across Task Hierarchies in Task-based Parallel Programming Frameworks
  • May 2025[c19][IPDPS'25 (opens in new tab)]Optimizing Fine-Grained Parallelism Through Dynamic Load Balancing on Multi-Socket Many-Core Systems
  • Jan 2025[i2][Preprint (opens in new tab)]MOFA: Discovering Materials for Carbon Capture with a GenAI-and Simulation-Based Workflow
  • Nov 2024[c18][FTXS'24 (opens in new tab)]Octopus: Experiences with a Hybrid Event-Driven Architecture for Distributed Scientific Computing
  • Sep 2024[c17][NRDPISI-1 (opens in new tab)]Diaspora: Resilience-Enabling Services for Real-Time Distributed Workflows
  • Sep 2024[c15][eScience'24 (opens in new tab)]TaPS: A Performance Evaluation Suite for Task-based Execution Frameworks(eScience 2024 Best Paper)

Cloud Computing

Distributed Systems