I'm teaching a performance engineering class at UC Davis (EEC 289Q) that's based on the amazing 6.172 performance engineering class from MIT. While many instructors have taught this class, Charles Leiserson has been the principal instructor for many years, and it's been a pleasure working with him and his colleagues in bringing it to UC Davis.

Much of the class has been a little more abstract, high-level, academic and I'd like to give a lecture that has several interesting practical examples that underscore some of the ideas that were taught in class. I've gathered a number of links from Hacker News over the past few months that are possible examples of what I'd love to hear (see below). The hope is that you might be able to suggest material or links that are along the lines of an in-depth blog post, hopefully with some interesting quantitative results, that basically tells a story like "I had this performance problem; it wasn't obvious that the answer to the problem was X, but I did some digging and it turns out it was X and I addressed it by Y." Ideally, X would be related to a topic we discussed in the class, which included:

General C-level performance optimizations / bit hacks / LLVM IR / compiler optimizations / races and parallelism / anything having to do with Cilk, which we studied in depth / parallel algorithms / storage allocation / cache {efficient, oblivious} algorithms / synchronization

Here's some of the links I collected from Hacker News that I thought might be suitable for ~10-15 minutes each:

https://hackernoon.com/timsort-the-fastest-sorting-algorithm-youve-never-heard-of-36b28417f399

https://github.com/apankrat/notes/tree/master/fast-case-conversion

https://github.com/madler/crcany

https://dlang.org/blog/2020/05/14/lomutos-comeback/

https://aqjune.github.io/posts/2021-10-4.the-select-story.html

https://nee.lv/2021/02/28/How-I-cut-GTA-Online-loading-times-by-70/

I'd be much obliged for any suggestions you might have!

user:	holla_a_m_ego
created:	November 25, 2021
karma:	2