[K12OSN] Anyone quantified savings using LTSP?
Terrell Prude' Jr.
microman at cmosnetworks.com
Fri Feb 8 04:29:43 UTC 2008
Rob Owens wrote:
> Terrell Prudé Jr. wrote:
>> James P. Kinney III wrote:
>>>> Of course, it's not that hard to come up with donated PCs, so that's a
>>>> bigger savings.
>>>>
>>>
>>> Old PC's are the way to get the thin client process _started_. The are
>>> large, hot, heavy and often come with the old CRT monitor. All power
>>> hogs.
>>>
>>
>> I have to disagree somewhat here. Old PC's actually don't use much
>> juice at all if you just pull the hard disks (the power cable to the
>> HD is sufficient) and use LCD screens with them. CRT's are going to
>> suck power whether you hook them up to a new Northtec-style thin
>> client or an old PC.
>> We once got a bunch of donated 15" LCD screens at one school, and
>> they got put into a lab full of Dell OptiPlex GX1's. Once I flashed
>> the BIOSes to the latest revision (took 5 minutes per box), things
>> went smoothly as wet ice...and saved a nice wad of cash, too.
>>
> I recently did some power measurements of various types of computer
> equipment. I'm attaching an ods spreadsheet, but I'm not sure if the
> list will block it or not. I found that old P2 or P3 machines used
> about 30 Watts or so (idle) when used as a thin client.
>
> -Rob
Looks like it came through OK.
Good, thorough analysis, Rob. Your measurements are similar to mine for
old PC's. I got 29W usage with a Dell OptiPlex GX1, with (IIRC) a
Pentium II-266MHz, HD disconnected. That is compared to my Athlon 2GHz
which will suck up as much as 150W if I run it full-tilt (I had the HD
doing a lot of work, too). I hadn't tried a thin client, since I don't
own one, but the power usage figures that you're showing for those are
impressive.
I guess it comes down to a cost-benefit analysis of power saved over the
expected remaining life of the old PC's whether you go w/ new thin
clients or old PC's.
--TP
More information about the K12OSN
mailing list