need for speed

I''m writting currently a 3D engine (or better I try to..) I want to speed up my math algorithms (matrix rotation, mul..) with inline assembler but the last time I programmed with this language was almost 4 years ago. So I hope you can tell me where I can find well commented source or tutos that could help me PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE (