Launching in 6 days: Vizuara's Kernel Engineering Workshop

GPU kernels are easier to understand when you can see what the hardware is doing. That's the idea behind Vizuara's Kernel Engineering Workshop, starting October 12, 2026. Explore the curriculum and register: https://kernelworkshop.vizuara.ai/ In this launch video, I share why I designed this workshop around visual explanations: following how data moves through a GPU, understanding streaming multiprocessors and memory, and connecting the code you write to the hardware running it. Across 25 live sessions, we'll pair concepts with coding and build from the foundations to modern GPU kernel engineering. The curriculum covers CUDA, memory optimization, matrix multiplication, tensor cores, profiling, FlashAttention, Triton, CUTLASS/CuTe, Hopper, Blackwell, and AI-written kernels. I'll be teaching alongside Shubham Panchal, with theory, live coding, and capstone project ideas from our workshop partner, Crusoe. You'll also get the illustrated Kernel Engineering book, session recordings, code files, and interview-preparation resources. If you want to understand GPU performance and build the skills used in modern AI and inference engineering, I'd love to see you in the workshop. #KernelEngineering #GPU #CUDA #Vizuara