As far as I know, the size of XMMATRIX must be 64 bytes, which is way too big to be returned by a function. However, DirectXMath functions do return this struct. I suppose this has something to do with the SIMD optimization. Should I return this huge struct from my own functions or should I pass it by a reference or pointer?

This question will look silly to you if you know how SIMD works, but I don't.

The DirectXMath library is inline, so no copy will happen on return, the function is just expanded in place.

Thanks.

DirectXMath uses the vectorcall calling convention that's available in VC++, which allows returning values in XMM0 through XMM3. This is enough to fit a 4x4 matrix, so even if the functions aren't inlined the return value doesn't have to go on the stack.