P3 · Computer Science & Computer Architecture
algorithms & data structures, how the machine executes
What a computer actually does with your code: how it represents data, why GPUs parallelise, and what a data centre is — the mental model that turns 'magic' into machinery.
This pillar opens the machine itself. Every other chapter on the platform talks about computers — models that need "more memory", training that needs "a GPU cluster", assistants that answer from "the cloud" — and this is the one place that explains what those words physically mean. No code, no electronics, no maths: each idea arrives as a picture or an everyday analogy, like the professional kitchen that stands in for a computer's three core parts.
The route is short and deliberate: first one ordinary computer and how its processor, memory and storage share the work; then why a graphics card's thousands of simple cores suit AI arithmetic; finally why the largest models outgrow any single machine and move into data centres. Climb those three rungs and headlines about chip shortages and gigawatt server farms stop being magic and start being machinery.