Updated On - 03:09:58 PM CDT November 04, 2022

Research and Experience

I’m a Computer Science Ph.D. Student at Northwestern University. I’m part of the Prescience Lab at Northwestern, led by Peter Dinda. Broadly, my interests lie in the realm of high-performance computing, extremely heterogeneous systems, and performance portability. More specifically, I’m interested in improving the utilization of diverse hardware resources within/across systems by looking at solutions bridging layers of the hardware-software stack.

You can download my CV here.

Education

Northwestern University | September 2021 - Present

I am currently a second-year Ph.D. student on the Systems track in Computer Science at Northwestern University. I’m interested in extreme heterogeneity as it relates to HPC systems.

Clemson University | August 2016 - May 2021

I graduated in May of 2021 with a degree in Computer Engineering from Clemson University. Some of the courses I have taken include: HPC Fault Tolerance, Digital Design using VHDL, Computer Organization, Embedded Computing, GPU Programming, and Operating Systems. In addition to my degree, I also obtained a minor in Spanish Studies.

Research

Northwestern University | PhD Student | September 2021 - Present

I’m currently a second-year PhD student, part of the Prescience Lab at Northwestern University. My research focuses on addressing challenges related to extreme heterogeneity in future parallel, distributed, and HPC systems through HW/SW Co-design of programming languages, compilers/runtimes, and architectures.

Sandia National Laboratory | R&D Intern - Grad | June 2022 - September 2022

Over the summer of 2022 I worked as an R&D Intern in the Scalable Algorithms Group at Sandia National Laboratory. My work involved evaluating novel architectures for viability in future heterogeneous systems. I focused on Nextsilicon’s HPC focused architecture and toolchain, as well as it’s interaction with the Kokkos programming model for performance portability.

Clemson University | Research Assistant | May 2020 - October 2021

Researched the benefit of SIMD parallelism on lossy compression for CPU architectures using leading lossy compressor for scientific data, SZ.
Investigated performance impact of alternative block padding methods on compression ratio and data quality of scientific data.
Contributed to larger collaborative projects involving groups from Clemson University, Washington State University and Argonne National Laboratory.
Studied a suite of common HPC workloads under a variety of domains (HACC, CESM, QMCPack, etc) to use in testing compression performance.

Clemson University | Creative Inquiry | January 2019 - May 2021

Studied a suite of common HPC workloads under a variety of domains (HACC, CESM, QMCPack, etc) to use in testing compression performance.
Investigated structure and performance characteristics of extreme scale systems by designing a system modeled after a Top500 supercomputing cluster.
Studied performance modeling of large scale systems using MPI, OpenMP and common HPC benchmarks like HPL and HPCG.

Publications

Griffin Dube, Jiannan Tian, Sheng Di, Dingwen Tao, Jon Calhoun, and Franck Cappello, “Efficient Error-Bounded Lossy Compression for CPU Architectures,” To Appear in 2022 30th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS).
Griffin Dube, Cavender Holt, John Hollowell, Sarah Placke, Sansriti Ranjan, Nikolas Heitzig, and Jon Calhoun, “Critique of “MemXCT: Memory-Centric X-Ray CT Reconstruction With Massive Parallelization” by SCC Team From Clemson University," in IEEE Transactions on Parallel and Distributed Systems, vol. 33, no. 9, pp. 2054-2057, 1 Sept. 2022, doi: 10.1109/TPDS.2021.3108961.

Work Experience

Delta Air Lines | Simulator Engineering Co-op | August 2018 - May 2020

Performed in depth research and modification of simulated aircraft systems and flight models for Boeing aircraft.
Streamlined calculation of balanced field length for B777 aircraft by designing MATLAB scripts which can be reused for all fleets, saving 1000+ hours of labor.
Adhered to FAA National Simulator Program regulations by documenting procedures used to correct issues found in software.
Communicated updates in simulator software to customers in order to receive feedback on operation.

Southern Machine Services, LLC | Mechanical Drafting & Design | May 2016 - August 2016

Created/designed various machine parts based on customer specifications using SolidWorks CAD software.
Project lead for design of safety housing for machine to cut steel cable based on customer specifications.