HPC Library for Exascale Linear Solver
HPC Library for Exascale Linear Solver
November 1, 2025
Project Duration: September 2025 ~ August 2027
Total Grant Amount: KRW ~200,000,000 (approx. USD 200K)
Project Overview
• Develops a scalable, high-performance numerical library targeting sparse and dense linear solvers for exascale supercomputers
• Designs six specialized solver modules including tridiagonal matrix solvers, Poisson solvers, and unstructured sparse matrix solvers (e.g., PaScaL_TDMA, UCFD_SPARSE, OCTREE_POISSON)
• Implements CPU/GPU heterogeneous parallelism, vectorization, and distributed computing optimizations
• Validates performance on national supercomputers (e.g., Nurion 5th/6th gen) with superior scalability (up to 120 PF), outperforming competing libraries like STRUMPACK and SuperLU
• Engages with domain researchers across academia and industry for use cases in CFD, electromagnetics, and fluid-structure interaction simulations
Join Us
We seek motivated students and researchers who want to grow as independent scholars and contribute to high-impact data and AI-centric research.
M.S. / Ph.D. Students
We welcome motivated graduate students who want to grow as independent researchers. Students are expected to identify important problems, design rigorous solutions, conduct careful experiments, and communicate their work through high-quality publications.
- Develop research taste through paper reading, discussion, and problem formulation
- Design and evaluate data and AI-driven approaches, and publish at leading venues
- Collaborate with academic and industry partners on impactful research
Postdoctoral Researchers
We welcome postdoctoral researchers interested in leading data and AI-centric research with strong academic and real-world impact. You will have opportunities to define independent research directions, collaborate with international partners, and contribute to high-impact publications.
- Independent research leadership
- International and industry collaboration
- Publication-driven academic career development
Undergraduate Interns
Undergraduate students can join research projects and develop hands-on experience in data and AI research.
- Research-oriented internship opportunities
- Mentoring from graduate students and faculty
- Pathway to advanced graduate research
Please email your CV, transcript, and brief research interests to Prof. Kwanghyun Park.