On Tue, Nov 30, 2010 at 11:59:57PM -0800, Tom Perrine spake thusly: > (This is why HP and some other blade chassis actually shut down extra > power supplies when the internal load is low, it is better to run > fewer supplies are closer to 90-100% of their max, than to run more > supplies at 20-30% of the rated loads.)
That is surprising to me. I thought running a power supply at 90-100% load would result in the faster failure of the power supply? Especially since they are nearly all over-rated. Maybe this is only good advice for properly constructed HP blade chassis power supplies and not for typical servers. Anyone have any opinions on PSUs with power factor correction (PFC)? Some good links: http://en.wikipedia.org/wiki/Switched-mode_power_supply#Power_factor http://en.wikipedia.org/wiki/Power_factor http://home.earthlink.net/~jimlux/hv/pfc.htm It seems that by using PFC power supplies in a large datacenter one could draw less current In my datacenter in San Diego, CA power (cooling is factored into the price also) is the most expensive thing I buy. Followed by bandwidth followed by physical space. It seems that if I were to use PFC in my power supplies I might be able to save some real money. I haven't gotten around to doing the tests though. Most people don't know what I'm talking about when I ask if their PSU has power factor correction so although I've been looking into it for nearly 4 years now I still don't have many power supplies with PFC. Still wishing the Google mobo's with built-in battery etc. would become a mass produced industry standard too. There are much better ways to do what we do but because of economies of scale and massive inertia in sticking with the old standards we are stuck. -- Tracy Reed http://tracyreed.org
pgphlQOqlSHwq.pgp
Description: PGP signature
_______________________________________________ Tech mailing list [email protected] https://lists.lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
