HPC Systems and Libraries

High-Performance Computing (HPC) systems and libraries have been playing a critical role for science discovery over the last decades. Recent years, HPC technologies have been heavily used by machine learning and deep learning applications. Advances in HPC systems and libraries are enabling researchers to solve bigger and harder problems.

(Note: Logo is designed by PADSYS Lab. All rights reserved.)

PADSYS Lab members design and develop novel HPC systems and libraries for both scientific computing applications and artificial intelligence workloads. We are particularly interested in the following topics:

• Message Passing Interface (MPI) Runtime
• Checkpoint-Restart Schemes
• Partitioned Global Address Space (PGAS) Runtime
• Hybrid MPI+X Programming Models

Acknowledgement

Thanks a lot for NSF's support on this research direction!

Grant Information:

• BD Spokes: SPOKE: MIDWEST: Collaborative:Advanced Computational Neuroscience Network (ACNN)

Software Release

• NVMe-CR. Will come soon!

• MVAPICH: MPI over InfiniBand, Omni-Path, Ethernet/iWARP, and RoCE. Prof. Xiaoyi Lu was a core member in the MVAPICH team when he worked at NOWLAB@OSU.

• High-Performance Neuroscience (NeuroHPC). Prof. Xiaoyi Lu was the co-founder and R&D leader for this project when he worked at NOWLAB@OSU.

• DataMPI: Extending MPI for Big Data with Key-Value based Communication

More information