BrrrViz Information

The interactive curriculum for understanding how AI runs on hardware. Most engineers who work with AI have never seen inside the machine. They use PyTorch, they call APIs, they tune hyperparameters, but the silicon beneath it all is a black box. BrrrViz closes that gap. We build interactive visualizations that take you from "I use ML" to "I understand what's happening on the GPU." Not documentation. Not textbooks. An experience where you can see threads execute, watch memory move, and build the mental models that make you a better engineer. Act 1: GPU Programming The machine itself. Execution model, memory hierarchy, occupancy, warp divergence, shared memory. The fundamentals that everything else builds on. Act 2: ML Systems (coming soon) How transformers map to hardware. Flash Attention, KV cache, inference bottlenecks, multi-GPU parallelism. The layer between "I train models" and "I understand why training is slow." Built for those who learn better by seeing.