If you would like to learn more about basic performance techniques and processor architectures, these are the resources I'd recommend:
- The short book, The Performance of Open Source Applications, edited by Travish Armstrong (available on http://aosabook.org, 2013) contains case studies showing mostly applications of the basic principles we discussed, for example, in the chapter about the Ninjia build system.
- Books about C++ performance, such as, this admittedly somewhat older books, Efficient C++ Performance Programming Techniques, by Dov Bulka and David Mayhew, Addison Wesley 1999, (by the way, this was my first computer performance book!) or a more recent Optimized C++ by Kurt Guntheroth, O'Reiily 2013, discuss many of the traditional basic performance techniques in their first chapters.
- Power and Performance. Software Analysis and Optimization, by Jim Kukunas, Morgam Kaufman 2015, discusses the history and architecture of modern Intel processors, introduces tools for obtaining CPU performance data and explains some low-level performance techniques. However, this book concentrates on Linux operating systems and tools.
- An even more in-depth discussion of performance gotchas in different processor architectures can be found on Agner's Fog site (http://www.agner.org/optimize/microarchitecture.pdf and others, 2018) and is directed more toward compiler writers. The widely cited article by Ullrich Drepper What every programmer needs to know about memory (available on http://www.akkadia.org/drepper/cpumemory.pdf, 2007) provides many details on RAM memory, caches, virtual memory, and possible low-level memory optimization techniques.
- Lastly, the most famous quote about performance optimization by Donald Knuth can be looked up in his article Structured programming with go to statements, ACM Computing. Survey., 6(4), 1974.
