Original Post
Hi.
I am going to write a efficient 4x4 matrix library. I am thinking of whether I can boost performance from the CPU''s power, eg. Pentium "SIMD/SSE" and AMD "3DNow!" --- (I just quote the names, I really don''t know the details...)
I have heard that using these technologies can greatly improve the speed of vector/matrix calculations. But I don''t know what''s the hell inside. So, my question is, can I just write my class using ordinary C or C++, and have the source code compiled by decent compilers that can automatically generate such optimized code?
You know, I didn''t want to learn the details of MMX/SIMD/whatever. I only know how to use Visual C++, and just wants a fast & easy way to make my executable runs faster - which cannot be achieved by normal C/C++.
Thx for any info.