0PSU LogoLow power PC — new or old PSU

I have a serv­er that oper­ates 24.7. It occurred to me that I may be able to save some money on elec­tri­city bills if I could reduce its power con­sump­tion. It already has a low power AMD Sem­pron LE-1100 CPU so the power draw of the sys­tem should be quite low, mais l' PSU is quite old and not espe­cially effi­cient. This got me think­ing about Power sup­plies designed for low power sys­tems. Il y a beaucoup de 80+ cer­ti­fied PSU« S là-bas avec 500+ watt rat­ings, but are these actu­ally worth the invest­ment for a low power system?

My sys­tem spends most of its time idle so I decided I’d work out the power draw would be if I were to build a new sys­tem with cur­rent com­pon­ents to do it’s job..

Base sys­tem (e.g. atome ou Core i5)18W
4x 2 To (e.g. Sam­sung Eco­Green F3)19W
RAID Carte (e.g. Areca 1220)5W
1x 120mm Case / HDD Ventilateur (e.g. Nex­us Real Silent 120mm)1W
TOTAL43W

Using the above fig­ures, I estim­ate that a new sys­tem with a sim­il­ar role and cap­ab­il­ity to what I have will con­sume just under 45watts at idle.

For a typ­ic­al low spec (450w) PSU, cette charge est seulement 10%. At this low a load a typ­ic­al PSU can be quite inef­fi­cient. le ATX spe­cific­a­tion requires 70% effi­ciency at 20% de la puissance nominale, mais un tirage au sort est inférieure à 43W 20% puissance nominale pour tous PSUEst évalué pour 250W ou au-dessus (en effet, tout PSU'S). Below is a com­par­is­on between a bare min­im­um ATX PSU and a few highly effi­cient PSU'S.

The fig­ure for the gen­er­ic PSU is based on the min­im­um require­ment and is there­fore a worst case. Even with a very poor effi­ciency PSU le tirage d'alimentation de la paroi ne sera pas augmenter à mesure que le tirage au sort DC diminue. I have cal­cu­lated a best-case yearly sav­ing, based on a kWh charge of 10p and assum­ing 24.7 usage.

Make / ModelEffi­ciency at 43WEst. tirage total de puissance (tourner au ralenti)Sav­ing
ATX v2.2 250W ATX70% (à 50W)61W
Season­ic X‑40080.4% (à 42.6W)53W6,97 £
Cor­sair AX85077.3% (à 43W)55W5,09 £
Ener­max Modu87+ 500w77.5% (à 43.4W)56W5,21 £
Cor­sair CX400W73.2% (à 43.2W)59W2,35 £
Sparkle Power SPI220LE Flex­ATX 220W80.5% (à 42.4w)53W7,02 £
HuntKey Jump­er R90 300W 80+ Gold85% (à 30w)51W£ 9.50

Con­clu­sion: For a low power sys­tem it is cheap­er, and prob­ably more envir­on­ment­ally sound to keep (ou réutilisation) an exist­ing power sup­ply than to replace it with a high power mod­el, even if the new mod­el is very efficient.

Mettre à jour (25-Dec-2011)

Cor­rec­ted a cal­cu­la­tion error and added the HuntKey Jump­er R90. Even a very good value, highly effi­cient PSU can­’t change the con­clu­sion much in this scen­ario — the sav­ings are just far too small to jus­ti­fy the cost (and envir­on­ment­al cost) d'une nouvelle PSU.  Le HuntKey prendrait plus de 3 années à payer pour elle-même, and this is in an ideal scen­ario, com­pared to a a worth case scen­ario PSUFor an even lower power sys­tem (e.g. AMD E350) the sav­ings would be lower still. Even a “per­fect” PSU would only net sav­ings of around £16 a year!

Laisser un commentaire