> 50u racks are also available (I've seen them in colo facilities)... > personally I don't like racking things that far over my head.
I don't really care about that, myself - racking any serious number is a two-person job anyway. but taller racks make some sense to me, since open space above the rack is pretty much wasted and/or a design flaw. (that is, you want your hot-air plenum to seal the top of the rack, which is most easily done with a drop-ceiling that meets the top.) density is an interesting question - if nodes are 350W, does it make sense to have 50 of them in a rack? 17.5 KW requires 5 tons of cooling, and as a rule of thumb 5 perf tiles. you can probably get by with 2-ish high-perf tiles if your air pressure is good. but that means your cold rows will be at least 4 tiles wide, possibly more. oops, wasn't density the argument for taller racks? > We've being switching cabinets in one of our datacenters to 220volt > service to support the sort of density we're seeing without running new > conductors. sure, there's no downside to 220 afaikt. one thing I don't fully understand is where the 80% figure comes from. just a fudge-factor, and what if you wind up always using that extra capacity? presumably an L6-30 circuit is actually safe to run at 30... _______________________________________________ Beowulf mailing list, Beowulf@beowulf.org To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
