Compute & Hardware
GPUs, Memory and FLOPs: Why AI Needs Them
Matrix maths, memory bandwidth and why inference is often memory-bound.
Intermediate 15 min For: Beginners / Developers / Researchers ✎ In progress
✎ This lesson is still being written. Below is the planned outline; it is not indexed by search engines yet.
What this lesson will cover
- Parallel matrix multiplication
- FLOPs vs memory bandwidth
- Unified memory on Apple silicon
- Estimating hardware needs