This is a great book for 3 reasons: 1) Abrash is actually a good writer. Sometimes, I read the beginnings of chapters just for the stories he tells. 2) It's ful

This is a great book for 3 reasons: 1) Abrash is actually a good writer. Sometimes, I read the beginnings of chapters just for the stories he tells. 2) It's full of algorithms. Sure, you probably won't be writing your own triangle rasterizer in assembly language, but the algorithms are still important to know. Besides, it has plenty of stuff that's pretty timeless (there is a boyer moore implementation in there). 3) Abrash's philosophy of optimization is the best there is.