◆ CIRCUITS & CHIPS, DECODED FOR THE AI ERA 10 FREE CALCULATORS NO LOGIN · NO ADS · NO TRACKING BUILT FOR EE STUDENTS WORLDWIDE v0.2 LIVE
◆ CIRCUITS & CHIPS, DECODED FOR THE AI ERA 10 FREE CALCULATORS NO LOGIN · NO ADS · NO TRACKING BUILT FOR EE STUDENTS WORLDWIDE v0.2 LIVE
/ glossary · 11 entries

The glossary.

Short explanations of the ideas underneath modern computing. Written for undergraduates who want to know what's actually happening on the silicon.

/ ENTRY 01
What's inside a GPU?
Streaming multiprocessors, CUDA cores, tensor cores.
/ ENTRY 02
What does "3nm" actually mean?
Why process node numbers are marketing.
/ ENTRY 03
Transistor to CPU — the stack
From one MOSFET to billions on a chip.
/ ENTRY 04
FPGA vs ASIC vs GPU
Three ways to run the same math.
/ ENTRY 05
Why power is the real bottleneck
Dennard scaling died. What replaced it.
/ ENTRY 06
Memory hierarchy: L1 to HBM
Speed vs capacity, always.
/ entry 07
What is a Tensor Core
The specialized matrix-multiply units inside NVIDIA GPUs that made modern AI possible.
/ entry 08
What's inside an H100
80 billion transistors, 132 SMs, HBM3 memory — the chip that made modern LLMs practical.
/ ENTRY 09
CPU vs GPU vs NPU vs TPU
Four processors, four philosophies, and why modern AI uses all of them together.
/ ENTRY 10
What is CUDA (and why NVIDIA won AI)
The 2004 software bet that quietly became a $3 trillion moat.
/ ENTRY 11
What is a MOSFET
The voltage-controlled switch that powers every modern chip — and the one datasheet detail that causes 90% of MOSFET failures.
expanding
More entries every week
Cache coherence, branch prediction, SIMD, tensor cores in detail, and more. Suggestions welcome at hello@circuitcopilot.com.