Power lines
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Carlos O'Donell carlos at baldric.uwo.caTue Jun 18 11:35:01 PDT 2002
- Previous message: Power lines
- Next message: gromacs benchmark and quad Xeons
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
> Hi, > We have a 72 nodes cluster. It seems this concentration of computers > is causing a distorsion of the power lines that could be the origin of failures > in other lines of the building. The problem is related with the > differential interrupters. > Is there a "correct" way to connect a cluster to the electrical network > or filter the noise that it generates ? > Thanks. > I'm glad I speak Castellano ;) By 'differential interruptor' do you mean 'differential switching supply'? Or possibly just 'switching supply'? If I remember correctly, this is a discussion that has come up before on the list. As most agreed, computer supplies have horrible harmonics and distort the power waveform usually at the peak of the cycle (flattening the peak). The only way to correct this is to hire a power-engineer to install harmonic correcting equipment (usually in the form of special power filters) or to do some type of harmonic analysis on the grid. I would be surprised if its the cluster that's behind the problem... but rather some large banks of power hungry UV Equipment :) c.
- Previous message: Power lines
- Next message: gromacs benchmark and quad Xeons
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
