[ntp:questions] is USB inherently evil, or only if something else is on the bus?

Rick Jones rick.jones2 at hp.com
Thu Aug 9 21:02:49 UTC 2012

David Taylor <david-taylor at blueyonder.co.uk.invalid> wrote:
> Yes, that's a good point, Rick.  Most PCs here are on a 1 Gb/s
> connection, and I've made no attempt to check whether 100 Mb/s would
> give better NTP performance.  For my own needs, what I get is
> already "good enough", but as soon as you start measuring you want
> to get the very best, even if it's not needed.....

My recollections may not fit well with unruh's data (or perhaps they
do), but interrupt coalescing is not completely correlated with
interface speed.  It is a feature of implementation, not media.  Now,
it is generally the case that interrupt coalescing appeared first in 1
GbE NICs, and so the prior 10/100 Mbit/s didn't have it, but that
doesn't necessarily mean that a 1 GbE NIC operating at 100 Mbit/s
won't still have interrupt coalescing going.

So I would think it isn't so much a matter of operating at 100 Mbit/s
vs 1 Gbit/s as it is the interrupt coalescing settings.


* of course since interrupt coalescing can be timing related and since
  operating at 100 Mbit/s will have different timing than 1 Gbit/s...

The glass is neither half-empty nor half-full. The glass has a leak.
The real question is "Can it be patched?"
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

More information about the questions mailing list