Larry Wigington
Lieutenant Colonel, United States Marine Corps
Last updated: April 2026
Summary
Senior military officer and operations research scientist with 20+ years of progressive leadership across law enforcement and the Marine Corps, currently completing a PhD focused on GPU-accelerated optimization for large-scale stochastic programs. Experienced bridging advanced computational research and executive-level decision-making across the Marine Corps warfighting enterprise. Proven ability to translate GPU computing and AI methods into operational capability at institutional scale.
Education
PhD, Operations Research
Naval Postgraduate School — Monterey, CA
Dissertation: Solving Large-Scale Stochastic Programs with High-Performance Distributed Computers
Master of Science, Operations Research
Naval Postgraduate School
Graduate Certificate, Operational Data Science & Statistical Machine Learning
Naval Postgraduate School
Graduate Certificate, High Performance Computing
Naval Postgraduate School
Master of Systems Analysis
Naval Postgraduate School
Bachelor of Science, Criminal Justice
Troy University
Professional Experience
PHDP-T Fellow
Naval Postgraduate School
Competitively selected for the Marine Corps Doctor of Philosophy Technical Program (PHDP-T) — a fully-funded research assignment under Marine Corps orders providing doctorate-level technical expertise in support of senior leader decision-making and long-range capability development. Conducting doctoral research in Operations Research focused on GPU-accelerated optimization for large-scale stochastic programs.
- ▸ Developing GPU-accelerated Splitting Conic Solver (SCS) exploiting block-angular structure of the two-stage stochastic program coefficient matrix via custom GPU kernels (CuPy RawKernel/RawModule)
- ▸ Presented research at ICSP 2025 (Paris) and SIAM Northern and Central California Sectional Conference (2025); accepted to present at SIAM OP26 (Edinburgh, June 2026)
- ▸ Dissertation applies GPU-SCS to novel portfolio optimization problems under uncertainty — including a stochastic programming formulation of Marine Corps investment portfolios — a problem class currently addressed by discrete scenario selection rather than full stochastic programming
Operations Research Analyst
Manpower and Reserve Affairs, Headquarters Marine Corps
Led quantitative analysis supporting manpower policy and personnel system decisions for the Marine Corps' 186,000-person active force.
- ▸ Designed and deployed a Python-based analytical tool for the $2.3 billion BAH appropriation program, reducing financial analyst processing time from 7+ days to under 10 minutes (99%+ reduction)
- ▸ Applied optimization and statistical modeling to officer talent management policy, directly influencing MOS restructuring for the 88XX/8825 technical officer community
- ▸ Delivered analytical products to SES- and flag-level decision makers at HQMC
Administrative and Operations Officer (Progressive Billets)
United States Marine Corps
Progressive operational leadership across tactical, administrative, and strategic billets. Led teams ranging from small tactical units to enterprise-level staff organizations. Managed complex multi-stakeholder operations under time pressure and resource constraint.
Technical Skills
GPU Computing
CUDA, CuPy (RawKernel/RawModule), Numba, Julia GPU kernels, NVIDIA RTX A3000/A6000, cuSolver, cuBLAS
Optimization Solvers
SCS, PDLP, ADMM, Benders Decomposition, Progressive Hedging, Pyomo, HiGHS
Languages
Python, Julia, R, MATLAB
Infrastructure
Linux (Ubuntu), Docker, GitLab CI/CD, Jupyter, LaTeX
Parallel Computing
GPU-native solvers, MPI, high-performance computing
Selected Research
Portfolio Optimization Under Uncertainty dissertation · in progress
Novel two-stage stochastic programming formulation for portfolio optimization under uncertainty — with application to Marine Corps investment portfolios (e.g., munitions procurement). Current DoD planning methods select from discrete scenarios; this work formulates the full stochastic program and evaluates GPU-SCS on the resulting instances.
GPU-Accelerated SCS for Two-Stage Stochastic Programs
Custom GPU kernels (CuPy RawKernel/RawModule) exploiting block-angular structure of the 2SSP coefficient matrix — parallel sparse operations for K, KT, and (I + KTK) across A, Ts, and Ws sub-matrices. Structure-aware CPU implementation exploits block-angular sparsity for near-peak memory bandwidth; GPU implementation operational with kernel tuning ongoing. Presented at ICSP 2025 (Paris) and SIAM NorCal 2025.
Convergence Analysis of SCS Enhancements for 2SSPs
Evaluated Ruiz normalization, adaptive/warped proximal operators, and Anderson acceleration on 2SSP instances (10–1,000 scenarios). Structure-aware solver benchmarked against standard SCS on SIPLIB instances; combined enhancements reduce iterations by up to 67%. CG iteration limiting is the dominant practical accelerator. 64-bit precision required — 32-bit arithmetic unstable.
Presentations
Wigington, L. "GPU-Accelerated SCS." SIAM Conference on Optimization (OP26), Edinburgh, Scotland.
Wigington, L. "Solving Stochastic Programs with GPUs: A Literature Review." International Conference on Stochastic Programming (ICSP), Paris, France.
Wigington, L. Poster presentation. SIAM Northern and Central California Sectional Conference.
Publications
Wigington, L. (2023). "Human Capital." Marine Corps Gazette, 107(11), 18–21.
Horner, D., Wigington, L., and Yoshida, R. (2021). "Red Cell Analysis of Mobile Networked Control System Supporting a Ground Force." Center for International Maritime Security.
View Article →Wigington, L. (2021). "Red Cell Analysis for Mobile Networked Control Systems." Master's Thesis, Department of Operations Research, Naval Postgraduate School.
Security Clearance
Top Secret / Sensitive Compartmented Information (TS-SCI)
Active