John K. Holmen
board

John K. Holmen

John

About

John is currently pursuing a Ph.D. in Computing: Scientific Computing under the guidance of Dr. Martin Berzins at the University of Utah's School of Computing and Scientific Computing and Imaging Institute.

His dissertation research focused on the exascale computing and software goals of Utah's Carbon Capture Multidisciplinary Simulation Center (CCMSC) and is on portable, scalable approaches for preparing large legacy codes and asynchronous many-task runtime systems for forthcoming exascale systems. Practical application of these approaches has been demonstrated at pre-exascale across systems including the NSF Frontera[13], DOE Lassen[11], DOE Summit[13], NSF Stampede 2[10][7], DOE Theta[8], and DOE Titan[8] using the open-source ~1.5M line Uintah Computational Framework, an asynchronous many-task runtime system for which he is a principal developer, and varying states of the resulting heterogeneous MPI+PPL task scheduling approach. This approach is demonstrated using the Kokkos C++ Performance Portability Library, a performance portability layer (PPL), and a related approach for indirectly adopting a performance portability layer in large legacy codes.

His current research continues CCSMC goals through the ALCF's Aurora Early Science Program and aims to port these approaches to the forthcoming exascale DOE Aurora system. Current research also explores further generalization of performance portability solutions for high performance computing through the combination of Uintah with a dataflow graph library, Hedgehog, in collaboration with NIST.

Prior to joining the University of Utah, he was a Development Engineer at R.W. Beckett Corporation, where he began employment as a Kettering University co-op. While there, his research focused on the design and development of PIC-based embedded systems for UL-certified safety-critical commercial and residential combustion products.

John earned his M.S.E. in Computer Engineering from Kettering University in 2012 and his B.S. in both Computer Engineering and Electrical Engineering from Kettering University in 2010.

He is an Eagle Scout and was a DOE/NNSA Predictive Science Academic Alliance Program (PSAAP) II graduate research assistant with Utah's CCMSC, 2014 University of Utah School of Computing Graduate Fellowship recipient, 2010 Kettering University President's Medalist, and Hotsey Caddie Scholarship recipient.

Research

John's primary research interests lie at the intersection between Electrical Engineering and Computer Science in Computer Engineering with a deep-rooted passion for understanding how to improve the use of arbitrary hardware and software stacks in computer systems and sharing practices and experiences with related communities. His experience ranges from pre-/post-acceptance major high performance computing systems featuring bleeding edge many-core architectures and pre-/post-production embedded systems for UL-certified safety-critical commercial and residential combustion products to cryptocurrency mining systems and video game systems.

His formal research interests lie in high performance computing, large-scale simulation, performance portability, performance optimization, many-core architectures, parallel programming, system engineering, and user experience.

Problems of particular interest to him involve understanding how to portably run well and scale across major high performance computing systems, evaluating current and emerging architectures, and helping to refine software stacks and best practices for emerging systems.

Papers

[13] J.K. Holmen, D. Sahasrabudhe, M. Berzins. "Porting Uintah to Heterogeneous Systems". In Proceedings of the Platform for Advanced Scientific Computing Conference (PASC). 2022.

[12] J.K. Holmen, D. Sahasrabudhe, M. Berzins, A. Bardakoff, T. J. Blattner, W. Keyrouz. "Uintah+Hedgehog: Combining Parallelism Models for End-to-End Large-Scale Simulation Performance". Refereed Paper for the Workshop on Hierarchical Parallelism for Exascale Computing (HiPar). 2021.

[11] J.K. Holmen, D. Sahasrabudhe, M. Berzins. "A Heterogeneous MPI+PPL Task Scheduling Approach for Asynchronous Many-Task Runtime Systems". In Proceedings of the Practice & Experience in Advanced Research Computing (PEARC) Conference Series. 2021. DOI: 10.1145/3437359.3465581. Oral Presenter.

[10] J.K. Holmen, B. Peterson, M. Berzins. "An Approach for Indirectly Adopting a Performance Portability Layer in Large Legacy Codes". In Proceedings of the International Workshop on Performance, Portability and Productivity in HPC (P3HPC). 2019. DOI: 10.1109/P3HPC49587.2019.00009. Oral Presenter.

[9] J.K. Holmen, B. Peterson, A. Humphrey, D. Sunderland, O.H. Diaz-Ibarra, J.N. Thornock, M. Berzins. "Portably Improving Uintah's Readiness for Exascale Systems Through the Use of Kokkos". Technical Report UUSCI-2019-001, SCI Institute. 2019.

[8] B. Peterson, A. Humphrey, J.K. Holmen, T. Harman, M. Berzins, D. Sunderland, H.C. Edwards. "Demonstrating GPU Code Portability and Scalability for Radiative Heat Transfer Computations". In Journal of Computational Science. 2018. DOI: 10.1016/j.jocs.2018.06.005.

[7] J.K. Holmen, A. Humphrey, D. Sunderland, M. Berzins. "Improving Uintah's Scalability Through the Use of Portable Kokkos-Based Data Parallel Tasks". In Proceedings of the Practice & Experience in Advanced Research Computing (PEARC) Conference Series. 2017. DOI: 10.1145/3093338.3093388. Oral Presenter. Best Paper Nominee.

[6] Z.A. Bookey, J.K. Holmen, J.J. Hu. " Kokkos-Based Structured Grid Multigrid Solver".In Center for Computing Research at Sandia National Laboratories Summer Proceedings, Ed. J.B. Carleton, Ed. M.L. Parks. 2016.

[5] B. Peterson, N. Xiao, J.K. Holmen, S. Chaganti, A. Pakki, J. Schmidt, D. Sunderland, A. Humphrey, M. Berzins. "Developing Uintah's Runtime System for Forthcoming Architectures". Refereed Paper for the Workshop on Runtime Systems for Extreme Scale Programming Models and Architectures (RESPA). 2015.

[4] J.K. Holmen, D.L. Foster. "Accelerating Single Iteration Performance of CUDA-Based 3D Reaction-Diffusion Simulations". In International Journal of Parallel Programming. 2014. DOI: 10.1007/s10766-013-0251-z.

[3] A. Cermak, J.K. Holmen, M.H. Zadeh, J. Dargahi. "Virtual-Reality Rehabilitation Utilizing a Haptic-Enabled Hand Function Test". In Proceedings of the International Conference on Human-Computer Interaction. 2011. DOI: 10.2316/P.2011.747-032.

[2] J.K. Holmen, M.H. Zadeh. "Effects of Varying Haptic Feedback on Driver Distraction during Vehicular Window Adjustment". In Proceedings of the International Conference on Automotive User Interfaces and Interactive Vehicular Applications. 2010. DOI: 10.1145/1969773.1969789. Oral Presenter.

[1] D. Racine, J.K. Holmen, M.H. Zadeh, M. Thompson. "Improving Vehicular Window Control with Haptic and Visual Feedback". In Proceedings of the International Conference on Human Haptic Sensing and Touch Enabled Computer Applications. 2010. DOI: 10.1007/978-3-642-14064-8_55. Poster Presenter.

Theses

J.K. Holmen. "Portable, Scalable Approaches for Improving Asynchronous Many-Task Runtime Node Use". Ph.D. Thesis, School of Computing, University of Utah. 2022.

J.K Holmen. "Accelerating Single Iteration Performance of CUDA-Based 3D Reaction-Diffusion Simulations". M.S. Thesis, Department of Electrical & Computer Engineering, Kettering University. 2012.

J.K. Holmen. "Design & Development of a 120-240 VAC 50-60 Hz Precision Delay Timer". B.S. Thesis, Department of Electrical & Computer Engineering, Kettering University in collaboration with R.W. Becket Corporation. 2010.

Book Chapters

J.K. Holmen, A. Humphrey, M. Berzins. "Exploring Use of the Reserved Core". In High Performance Parallelism Pearls Volume Two: Multicore and Many-core Programming Approaches, J. Reinders, J. Jeffers, Ed. Morgan Kaufmann. 2015. DOI: 10.1016/b978-0-12-803819-2.00010-0. Poster Presenter.

Talks

"Using MPI+Kokkos to Portably Extend Uintah to Many-Core Systems". Presented at the Intel Extreme Performance Users Group Annual Fall Conference. 2018.

"GPU-Accelerated 3-Dimensional Reaction Diffusion Systems". Presented at the UIUC Beckman Institute for Advanced Science and Technology Workshop on GPU Programming for Molecular Modeling. 2013.

Patents

J.K. Holmen, M.H. Zadeh, M. Thompson. "Vehicular Window Adjustment by Means of a Haptic-Enabled Rotary Control Knob". U.S. Patent No. 8,786,417. 2014.