[time-nuts] [LEAPSECS] Leap second to be introduced at midnight UTC December 31 this year
michaeljwouters at gmail.com
Mon Jul 25 18:09:15 EDT 2016
I'd be interested to know how 10ns less latency is useful. AFAIK, a high
speed trading system consists of multiple processing nodes, which certainly
could have a 10ns accurate time reference eg via PTP, but latency and
jitter would limit time stamping of transactions to the microsecond level.
Unless of course, the critical bits are all running on dedicated hardware.
On Tuesday, 26 July 2016, bownes <bownes at gmail.com> wrote:
> A second here or there is a very big deal to those of us in the financial
> and database worlds.
> Aside from the well known instances involving electronic trading I have
> customers fighting over cabinet positions and cable lengths to place
> processors closer to disk drives and on switch paths with 10ns less latency.
> > Apple's new file system timestamps files with nanosecond resolution. A
> lot of Linux file systems also do that now. The nanosecond ain't what it
> used to be... I can imagine people wanting picosecond timestamps in the
> near future. Who knows, maybe we'll have something like NTP compensating
> for light-speed delay for the gap between the read/write heads and disk
> surface (assuming we still use heads and surfaces) ; -)
> > --------------------
> > Except that you still have the issue of “what time did this file get
> changed”. To most of us, that’s not really a big deal. A second either way
> … who cares.
> > _______________________________________________
> > To unsubscribe, go to
> > and follow the instructions there.
> To unsubscribe, go to
> and follow the instructions there.
More information about the time-nuts