Topic: “Structural and parametric mapping of Population-Based Metaheuristic Optimization Algorithms with GPU architecture”.
Defended PhD on Engineering on 15.12.2024 in Bauman State Technical University. Dissertation board info.
During the work on PhD, I designed an auto-tuning parametric optimisation platform for the optimal structural mapping of algorithms to the GPU. Developed custom LLVM-based compiler for transforming source code to GPU assembly, algorithms for adaptive control and transforming task interaction graphs. I have performed research on parallel and distributed computing, graphical processors, meta-heuristic algorithms, and parametric optimisation, resulting in more than 15 papers and conference talks.
I have authored and taught five courses on CUDA programming, machine learning (PyTorch, DNN, CNN, image recognition), LaTeX lections and workshops for graduate and postgrad students. Pioneered the GPGPU research at the department. Many new scientists currently work in this field at the department.
Slides and materials for my courses and workshops at Bauman Moscow State University.
Course “GPU programming with CUDA”, Fall 2011
Course “GPU programming with CUDA”, Fall 2012
Course “GPU programming with CUDA”, Fall 2013
Lection in “Parallel Computing”
Course “Machine Learning”, Fall 2022