[time-nuts] Timing performance of servers

[Magnus Danielson]

On 10/25/2012 01:17 PM, Bob Camp wrote:
> Hi
>
> Judah Levine (probably spelled his name wrong) from NIST has a series of papers on this. They go back into the 90's.

For once you got his name right :)

I will go back to his papers (NIST has 106 papers with his name on it) 
as there is surely a lot of things that he written that can be useful.

However, I wonder if he ever bothered to illustrate the issues that I 
wanted to educate folks with.

Judah showed the NIST time clock labs for us. They are now up to 386 
based machines to maintain the NIST time-scales. They only do work every 
12 min anyway, so it doesn't really care if they can cut time from 4 s 
to 0,4 s.

Cheers,
Magnus