> That's pretty neat. Is it actually any faster than just doing four 8-bit adds, though? The operation itself is not necessarily faster (it could be faster due

> That's pretty neat. Is it actually any faster than just doing four 8-bit adds, though? The operation itself is not necessarily faster (it could be faster due to pipelining, I think, and it would probably ne faster when storing 8 8-bit ints in a 64-bit int), but it can save the hassle and runtime of packing and unpacking into four variables.