1
CompaniesChapter 1 of 7
TSMC
Where most AI chips are manufactured.
TL;DR
TSMC is a chip foundry company based in Hsinchu, Taiwan. The world's largest semiconductor foundry.
Learning path7 terms · ~21 min read
From Sand to Model
The AI supply chain in 7 terms · foundry, memory, chip, system, training, inference, pricing.
Read full chapter
2
MemoryChapter 2 of 7
HBM3e
High bandwidth memory, the throughput backbone.
TL;DR
HBM3e (High Bandwidth Memory 3e (Enhanced)) is a JEDEC HBM memory generation from 2024, with 1,180 GB/s of bandwidth per stack.
3
ChipsChapter 3 of 7
GPU
Where the silicon becomes an accelerator.
TL;DR
A GPU is the accelerator that trains and serves most large AI models.
“A GPU is the accelerator that trains and serves most large AI models.”
Read full chapter4
SystemsChapter 4 of 7
DGX GB200 NVL72
Many chips wired into one rack-scale system.
TL;DR
DGX GB200 NVL72 is a NVIDIA AI system with 72 NVIDIA B200 and 36 NVIDIA Grace.
5
ConceptsChapter 5 of 7
Pretraining
What these systems are used for first.
TL;DR
The phase where a model learns from internet-scale text via next-token prediction · typically 15-30T tokens and millions of GPU hours.
“Pretraining is the moat.”
Read full chapter6
ConceptsChapter 6 of 7
Inference
What they are used for every day after that.
TL;DR
The process of running a trained model to generate predictions · every API call is inference.
“Inference optimization is where the next 10× cost reduction lives. Every frontier lab is racing to ship the best serving stack.”
Read full chapter7
PricingChapter 7 of 7
Input Tokens
How model compute finally becomes a price.
TL;DR
The tokens in your prompt · billed per million, typically 3-5× cheaper than output tokens.
“Input token discipline separates teams that can scale from teams that can't. Cache aggressively.”
Read full chapterWhat you learned
By the end you understand the full stack from silicon to inference API · and why each layer sets pricing for the one above it.
Keep learning
Next path · 7 terms
The AI Bubble Explained
Seven terms that decode whether AI is overpriced, fairly priced, or criminally underpriced. Read in order.