01Why GPUs Exist

Why GPUs Exist

FoundationsModule 014 topics~2h 30m

Topics, in order

01 What Is a GPU?A GPU (graphics processing unit) is a processor built to do the same simple calculation on a very large amount of data at once. A CPU has a few powerful cores that each finish one task … Beginner 30 min 02 Latency vs ThroughputChip designers get a fixed budget of transistors and watts. A CPU spends that budget on making one instruction stream finish as soon as possible (low latency). A GPU spends it on doing as … Beginner 40 min 03 From Pixels to TensorsGPUs were invented to draw video-game frames. Drawing a frame means running the same small calculation for millions of pixels, sixty times a second. Neural networks turned out to need … Beginner 35 min 04 Reading a Spec SheetA GPU data sheet lists dozens of numbers. For compute work, five decide almost everything: memory capacity, memory bandwidth, peak arithmetic rate (for the number format you will use), … Beginner 45 min

About this module

A GPU is not a faster CPU. It is a different bargain: give up speed on any single task, and in exchange do an enormous number of simple tasks at the same time.

This module builds that one idea from four directions.

#	Lesson	The question it answers
01	What Is a GPU?	What is physically in the box, and what does it do?
02	Latency vs Throughput	Why are CPUs and GPUs built so differently?
03	From Pixels to Tensors	How did a graphics card become the engine of AI?
04	Reading a Spec Sheet	Which numbers on a GPU data sheet actually matter?

When you finish you can explain to a colleague, without notes, why a GPU with 17,000 “slow” lanes beats a CPU with 64 fast cores on matrix multiplication — and loses on almost everything else.