Research Overview
My research focuses on understanding and optimizing the performance of high-performance computing systems, with a particular emphasis on I/O and scheduling.
Parallel I/O in HPC
I work on characterizing and improving the I/O behavior of HPC applications. This includes analyzing temporal I/O patterns across large-scale systems, developing bandwidth management strategies (such as IO-SETS), and studying the impact of storage configuration on application performance. A key part of this work involves collecting and analyzing I/O traces from multiple HPC platforms.
Key publications:
- A Deep Look into the Temporal I/O Behavior of HPC Applications (IPDPS 2025)
- IO-SETS: Simple and efficient approaches for I/O bandwidth management (IEEE TPDS 2023)
- The role of storage target allocation in applications’ I/O performance with BeeGFS (IEEE Cluster 2022)
Scheduling and Cloud Computing
My Ph.D. research addressed the challenge of scheduling bag-of-tasks applications on cloud platforms using spot (preemptible) instances. I developed dynamic schedulers that tolerate hibernations and leverage burstable virtual machines to reduce costs while meeting deadlines. This work has applications in bioinformatics, reservoir simulation, and other compute-intensive scientific domains.
Key publications:
- Scheduling Bag-of-Tasks in Clouds using Spot and Burstable Virtual Machines (IEEE TCC 2021)
- Comparing SARS-CoV-2 Sequences using a Commercial Cloud with a Spot Instance Based Dynamic Scheduler (IEEE/ACM CCGrid 2021)
- A dynamic task scheduler tolerant to multiple hibernations in cloud environments (Cluster Computing 2020)
Benchmarking and Performance Evaluation
I am the developer of IOPS (Integrated Orchestration for Parametric Studies), a generic benchmark orchestration framework that automates parametric experiments on HPC systems.