6 May
2010
6 May
'10
3:35 p.m.
On Thu, May 6, 2010 at 5:31 PM, Fred Cisin <cisin at xenosoft.com> wrote:
Bah, I say just go back to non-microprocessor based discrete logic consoles. ;)
Marty
> > I
still stand my my statement that the master clock rate is meaningless.
> Laws of physics disagree with you for sufficiently high orders of magnitude.
On Thu, 6 May 2010, Tony Duell wrote:
Care to explain? I can't think of a single
law of physics that prohibits
using as many clock cylces as I want for a particular operaiton. In which
case, an 8MHz RISC processor which takes 1 cycle for a given operation
could be a lot faster than a 1GHz processor what happens to take 256
cycles for the same operation.
. . . and, of course, at a higher level, any benchmark becomes a software
issue. ?It is easy to "prove" that a 4MHz Z80 is faster than a 4GHz
Pentium, just by setting a task, and stipulating that the task must be
accomplished using software that was/is commercially available.