Inside a Transformer
Follow inputs through attention and MLPs to the next token. Change the numbers and see what happens.
Not started
Learn how AI compute works.
Models, hardware and infrastructure. Hands-on courses and labs you can explore in your browser.
Courses
6 courses · progress saved in this browserGPU Kernels & Model Bring-Up
Write and optimize kernels in your browser, then move to CUDA and model inference.
Not started
How Data Centers Work
Follow power, cooling and networking from the substation to the rack, then learn how to choose a site.
Not started
Cooling & Powering an Inference Cluster
Trace a 5 MW cluster’s water and power from the systems to the plant.
Not started
Deploying a Cluster Network
Learn who orders, installs, configures and tests the network, from purchase order to burn-in.
Not started
Buying NVIDIA Systems
Understand OEMs, ODMs, allocation and the steps to getting racks delivered.
Not started
Interactive labs
4 spaces to exploreInside a GB300 NVL72
Pull out compute trays, reveal the chips and cooling, and follow a request through the rack.
Cluster Connectivity
Wire an inference cluster into a data center. Choose optics, connect handoffs and test redundancy.
Cooling & Power
Pipe and power an inference cluster, then test how the system handles failures.
Data Center Explorer
Walk the facility, open racks and inspect the hardware, cooling and electrical systems.