Using CUDA-Accelerated Binaries and Libraries in Julia with CUDA.jl 4.0

Learn how to leverage CUDA-accelerated libraries and applications in Julia through this 28-minute technical talk presented by JuliaHub software engineer Dr. Tim Besard. Explore the breaking changes introduced in CUDA.jl 4.0, including the complete overhaul of CUDA toolkit interactions, and discover how to adapt existing code to work with external CUDA-accelerated resources. Master the integration of CUDA-accelerated C applications and libraries, while gaining insights into recent improvements such as enhanced sparse array support. Understand the process of packaging GPU software for use with CUDA.jl, including proper handling of CUDA toolkit dependencies and practical implementation strategies. Build upon this knowledge by exploring GPU programming fundamentals through the complementary Julia GPU Programming webinar series.