[6160] in Release_7.7_team

home help back first fref pref prev next nref lref last post

Re: Putting display to sleep on Athena

daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Mon Jan 12 18:19:32 2009

Message-Id: <200901122318.n0CNIrVb006407@speaker-for-the-dead.mit.edu>
From: Jonathon Weiss <jweiss@MIT.EDU>
To: Laxmi Rao <lrao@MIT.EDU>
cc: Jonathon Weiss <jweiss@MIT.EDU>, Greg Hudson <ghudson@MIT.EDU>,
   release-team@MIT.EDU, William Cattey <wdc@MIT.EDU>,
   Jonathan McIndoe Hunt <jmhunt@MIT.EDU>
In-reply-to: Your message of "Mon, 12 Jan 2009 18:10:44 EST."
             <355EE540-695A-400B-9374-3402B935AC80@MIT.EDU> 
Date: Mon, 12 Jan 2009 18:18:53 -0500
X-Spam-Flag: NO
X-Spam-Score: 0.00



Bill pointed out that not every one of the 956 machines has a display
to sleep (eg, departmental server, or VM, or laptop that is off when
not in use), so that number is a little more aggressive than it would
otherwise be.  If you drop it to 900 machines, you still get $9k of
savings, and I really do think that's quite conservative, and we're
likely ot hit $10.  Unfortunatley, of course, we don't actually have
any way to measure it.

	Jonathon



> Jonathan,
>    It was goo dto see you today.
> The analysis is sound and conservative. We'll take the $10k in   
> annual savings
> Many Thanks
> Laxmi
> On Jan 12, 2009, at 6:00 PM, Jonathon Weiss wrote:
> 
> >
> > Laxmi,
> >
> > To followup on our hallway conversation this afternoon...  Below is
> > some of the resarch and details of the numbers we worked through to
> > estimate the cost savings of sleeping the displays on idle Athena
> > machines (both in clusters and private offices).  However the short
> > form is:
> >
> > Assume 25W/monitor savings (conservative)
> > Assume 1000 workstations (actually 956 according to Dec. stats)
> > Assume 10 sleeping hours per day (conservative and easy to multiply)
> > Assume 365 days per year :-)
> > Estimated power savings: 91.25MWh
> > Using your (conservative) blended average cost to MIT of $0.11/kWh
> > Estimated cost savings: $10037.05/year
> >
> > Of course, we really only have 1-2 significant digits, but we can
> > reasonably say $10K/year!
> >
> > If we assume 35W, and an average cost of $0.15 we're up ot $19K/year.
> >
> > 	Jonathon
> >
> >
> >> To follow up, I looked at the Dell and HP data sheets on the 3 common
> >> Athena video monitors for power consumption:
> >>
> >> This year's: Dell 2009W: Max: 35w, Minimum (displaying msdos prompt):
> >> 19.4w, Active-Off: .6w
> >> Last year's: Dell 2007WFP: Max: 75w, Typical (routine task): 55w, Off
> >> (still plugged in.): .85w
> >> Older HP 1702: Max: 40w, Typical: 28.5w, Power Save: 2w.
> >>
> >> I don't really want to do a weighted sum of monitors and power
> >> usage.  I think going with telling people "25 watts saved per monitor
> >> as a conservative estimate", or "roughly 100 MWH per year,
> >> conservatively", is the right thing.
> >>
> >> -wdc
> >>
> >>
> >>
> >> On Jul 29, 2008, at 4:38 PM, Jonathon Weiss wrote:
> >>
> >>>
> >>> This got me to thinking "this will probably save enough power to
> >>> actually be interesting, won't it?"  to answer that question for
> >>> myself and other's I present some numbers.
> >>>
> >>> Assume you can save 25W by sleeping an average Athena LCD monitor (I
> >>> beleive this is a conservative estimate, and the actual number  
> >>> will be
> >>> more like 30 or 35).
> >>>
> >>> Multiply by 1000 Athena workstations (Actually 984 in the July  
> >>> status
> >>> report, but 1000 is really easy to multiply): 25kW
> >>>
> >>> Assume that the average monitor will be sleeping 10 hours a day/ 
> >>> night
> >>> (I think this is very conservative, especially since it assumes
> >>> weekends are regular days, but again it is easy to multiply): 250kWH
> >>>
> >>> 365 days/year: 91.25 MWH/year.
> >>>
> >>> http://cogen.mit.edu/powermit/ estimates that N* bills MIT $58.77/ 
> >>> MWH
> >>> (though this number appears to be 5 years old, and is probably a
> >>> low-ball estimate now): $5363/year
> >>>
> >>> If we assume 35W savings and 14 hours a day of sleeping monitors
> >>> (which is certainly more agressive, but not unbelievable):
> >>> 490kWH/day
> >>> 178.85MWH/year
> >>> $10511/year
> >>>
> >>> Realistically, I'd guess that we'll fall between the two.  Either  
> >>> way,
> >>> it seems worth pursuing.
> >>>
> >>> 	Jonathon
> >>>
> >>
> >
> >
> 
> Laxmi Rao
> IT Energy Co-ordinator
> Information Services and Technology
> Massachusetts Institute of Technology
> Room N42-040f
> Cambridge, MA 02139
> 617-452-2333
> lrao@mit.edu


home help back first fref pref prev next nref lref last post