[Info-vax] AlphaServer DS10 470 MHz?

[JohnF]

In comp.os.vms VAXman- wrote:
> I suspect that it's far easier to accurately measure the clock speed than
> it is to try to measure the zero-crossings accurately for cycle-time.

Just curious -- I was wondering about that: isn't the cpu driven by
an (onchip or onboard?) oscillator of some sort? If that's the same
oscillator that drives the clock, then it's always going to measure
the "correct number" (i.e., however many cycles it's programmed to
count as a second) -- you can't know if your clock is running slow
by looking at that same clock.
-- 
John Forkosh  ( mailto:  j at f.com  where j=john and f=forkosh )