Attack at Dawn: Cracking Codes with CUDA and Teaching Parallelism

Workshop:EduHPC-21: Workshop on Education for High Performance Computing

Authors: Rahul Toppur and Michael Shah (Northeastern University, Khoury College of Computer Sciences)

Abstract: In this work, we present four introductory modules on General Purpose GPU programming (GPGPU) with a context in computer security for undergraduate students. Students work through a series of lectures, labs, and assignments learning parallel programming in CUDA while learning high-performance computing concepts including device memory management and the CUDA thread model. This course adopts a security focus to GPGPU programming and uses topics such as cracking ciphers and image steganography to provide students with a practical context for learning high-performance computing in a unique domain, which is insertable into a normal semester-long security course.

Back to EduHPC-21: Workshop on Education for High Performance Computing Archive Listing

Back to Full Workshop Archive Listing