A wrapper for intel SSE/AVX vector instructions
This is just a toy thing to figure out what working with intrinsics is like. I tried to keep it kinda short, which was possible with some defines and templating. Right now only basic operations are implemented (as far as possible), so no strange other AVX/SSE functions.
If you try to use an unimplemented / invalid operation
(for example multiplication on unsigned vectors is unimplemented
because no intrinsics exist, or shifting a floating point vector,
which is just invalid in general), it should compile to nullptr(), which
will cause the compiler to fail already.
I will probably add a matrix class later too.