The aim of this tutorial is to give non-professional programmers a brief overview of the features of modern hardware that you must understand in order to write fast code. It will be a distillation of what have learned the last few years. This tutorial will use Julia because it allows these relatively low-level considerations to be demonstrated easily in a high-level, interactive language.
https://viralinstruction.com/posts/hardware/
Table of contents:
- What this notebook is not
- Setting up this notebook
- The basic structure of computer hardware
- Avoid accessing disk too often
- Avoid cache misses
- Keep your data aligned to memory
- Digression: Assembly code
- Allocations and immutability
- Registers and SIMD
- Struct of arrays
- Specialized CPU instructions
- Inlining
- Unrolling
- Avoid unpredicable branches
- Be aware of memory dependencies
- Variable clock speed
- Multithreading
- GPUs
The site, viralinstruction.com, only has two articles, but the domain name sounds promising. 