On 13 Feb 2007 at 22:11, arcarlini at
iee.org wrote:
Quite how the VAX-11/780 became the reference 1 MIP
machine
while technically cranking through only half that many instructions
is something that I've never figured out. Anyone know?
Does a MIP have any practical meaning other than comparing two
machines of the same architecture?
I worked on a machine during the 70's that probably had an issue rate
comparable to the 11/780 but could crank out nearly 100 MFlops.
It reminds me of comparing clock speeds as a measure of dissimilar
processor performance. Wasn't there someone on the 6809 project who
said something to the effect that if clock speed really meant
something, they could have fit the 6809 with a waveguide?
Cheers,
Chuck