P08 · High-Performance and GPU Computing
A free, illustrated L3 course of 8 modules in the Programming and Data track. Modules include: Computer architecture for ML: CPU, memory and GPU; Profiling and vectorization; Parallelism: threads, processes and message passing.
P08 L3 · Advanced 8 modules · about 32 lessons Planned · Ring 4
Part of the Programming and Data track. Ships in Ring 4 (Expert) of the roadmap.
Modules
- Computer architecture for ML: CPU, memory and GPU
- Profiling and vectorization
- Parallelism: threads, processes and message passing
- The GPU programming model
- Tensors, strides and memory layout
- Automatic differentiation engines
- Just-in-time compilation and graph optimizers
- Mixed precision and memory management
Signature illustrations
The pictures a learner should still see in their head a year later:
Memory hierarchy pyramid, tensor stride and view diagrams, data-parallel versus task-parallel schematic, the autodiff tape.
Status
This course is planned, not written yet. It ships with Ring 4 (Expert).
Every lesson will follow the same twelve-part blueprint with at least three purpose-made figures. Watch the repository to hear when modules ship, or help write and review it.