[ntp:questions] NTP Performance on WINNT

David J Taylor david-taylor at blueyonder.not-this-part.nor-this.co.uk.invalid
Mon Jul 20 15:45:48 UTC 2009

paul wrote:
> Thank you, David.
> In my situation, no GPS is availiable. So can I expect better
> performance when GPS is used as reference clock, or when a stratum-1
> NTP server is added to the network?

Local GPS or a local stratum-1 server will make a significant difference. 
When using the Internet the quality of your connection to the remote 
server, and the remote server loading are probably more important factors 
than precisely how that server is synced (i.e. whether it has a local GPS 

> And, I noticed the offset for Narvik is more stable than that for
> Hydra, is it due the PPS feeded to Narvik? But Bacchus is doing good
> too without PPS, Why?

Hydra is running Windows-7.  Because Windows Vista and XP may have a 1KHz 
clock, the interpolation scheme which is used with XP does not work 
correctly.  So if you like to think of it this way, the Windows-7 system 
is using purely "native Windows 1KHz" timing, but the XP system is using 
"interpolated 64Hz" timing, and the guys who wrote and implemented that 
interpolation did a great job.  If Narvik and Hydra had the same OS, I 
would expect similar performance.

Bacchus does currently have a GPS/PPS signal, paralleled off one of the 
GPS receivers.  It runs Windows 2000.  It can be a bit of a testbed, so 
unlike Feenix and Stamsund the GPS/PPS is not guaranteed to be there. 
Without the GPS/PPS local reference clock, it performs similarly to 

We're into the region where the performance would be improved by keep the 
PCs, or at least their timing oscillator crystal, at a more constant 


More information about the questions mailing list