> > Removing all but the mov instruction from future iterations of the x86 architecture would have many advantages: the instruction format would be greatly simplified, the expensive decode unit would become much cheaper, and silicon currently used for complex functional units could be repurposed as even more cache. As long as someone else implements the compiler.
I realize it's tongue in cheek, but saying the decode unit in x86 is expensive is kind of a tired and untrue adage in today's computer architecture terms. But I guess it is fair to say that the die is stuffed with as much cache as they can fit :)
It may not be accurate to say "decode unit" anymore but who wouldn't welcome anything removing the large amounts of cruft from the ISA? Nobody uses most of the damn opcode space except where required for mode trickery or legacy code.
I realize it's tongue in cheek, but saying the decode unit in x86 is expensive is kind of a tired and untrue adage in today's computer architecture terms. But I guess it is fair to say that the die is stuffed with as much cache as they can fit :)