Subject: RE: VAX/VMS Time
To: 'Lennart Augustsson' <lennart@augustsson.net>
From: Antonio Carlini <arcarlini@iee.org>
List: port-vax
Date: 12/01/2003 20:23:09
> From: Lennart Augustsson
> adam.r.pigg@bnfl.com wrote:
> > Hi
> >=20
> > I was wondering if anyone could explain to me how a vax/vms system=20
> > (3100
> > m38) keeps track of time internally.
> >=20
> > I read on the internet that it uses a 64 bit number that=20
> counts in 100=20
> > nano second intervals starting from the 'Julian Day'  (Nov=20
> 17 1858) is=20
> > this correct?  If so how accurate is it because surely=20
> there would be=20
> > no way that the clock is this accurate!! :)
> Yes, it's counted in 100ns increments since the opening of=20
> the Smithsonian Institute (if memory serves me right).

The time is indeed maintained as a 64-bit quantity in 100ns
steps since 00:00 17-NOV-1858. However, VMS only requires
the system time to be updated every 10ms.

Various VAXen have interval counters which can count down
to 1us resolution (I think it is mostly the older pre-uVAX chip
machines which do this).

Antonio
=20
--=20

---------------
Antonio Carlini             arcarlini@iee.org