ECE 432GPU Programming

Southern Illinois University Carbondale · 3 credits

Catalog

This course provides an in-depth exploration of GPU architectures and parallel programming using CUDA C/C++. Students will study many-core programming priniciples, streaming multiprocessors (SMs), memory hierarchies, warp scheduling, and execution models that enable high-performance computing. The course covers CUDA programming fundamentals, including thread organization, memory coalescing, register optimizations, and efficient utilization of tensor cores for AI workloads. Advanced topics include cooperative thread programming, large-scale parallelism, and performance tuning using CUDA libraries such as cuDNN for deep learning and cuBLAS for optimized linear algebra operations. Architectural aspects such as cache optimizations, data locality, and tensor core acceleration will be analyzed for efficient GPU utilization. Prerequisites: ECE 321 and ECE 296 with grades of C or better. Lab fee: $20 to help defray cost of equipment.

Recent sections

Spring 2027 · 1 section

Jan 11 – May 7

SectionDays & timeLocationInstructor
001Class #25149Mo We Fr · 4:00 – 4:50 PMEngineering, E Wing 0136Iraklis Anagnostopoulos