Prerequisites

• Multivariable Calculus (Math 8)
• C(++) Programming (COSC 50).
• It would also be helpful to have prior experience with Linear Algebra/Optimization (Math 22 / COSC 70).
• Contact the instructor if you have equivalent prior experience but do not have prerequisites (e.g., pointers, arrays, and memory management)

Course Overview

Many stages of state-of-the-art robotics pipelines rely on the solutions of underlying optimization algorithms. Unfortunately, many of these approaches rely on simplifications and conservative approximations in order to reduce their computational complexity and support online operation. At the same time, parallelism has been used to significantly increase the throughput of computationally expensive algorithms across the field of computer science. And, with the widespread adoption of parallel computing platforms such as GPUs, it is natural to consider whether these architectures can benefit robotics researchers interested in solving computationally constrained problems online. This course will provide students with an introduction to both parallel programming on GPUs as well as numerical optimization. It will then dive into the intersection of those fields through case studies of recent state-of-the-art research and culminate in a team-based final project.

Learning Outcomes

By the end of the semester, you will be able to:

• Analyze the opportunities and limitations of parallel programming on GPUs
• Evaluate the strengths and weaknesses of numerical optimization algorithms
• Engage critically with recent research on parallel optimization algorithms for robotics
• Collaborate with a team to develop and present an open-ended final project

Office Hours

The most up-to-date schedule of office hours can be found here. I will also try to respond to requests emailed to plancher+courses@dartmouth.edu within 2 business days. Faster response time will be achieved via the course Slack.