It is also good to note that you may want to configure these to not use GPU's. Granted right now BOINC doesn't support Intel GPU's, future releases may. Supposedly the Sandy Bridge line and newer are capable once Berkeley gets around to it. WCG is still planning on having GPU support in the future and that will increase your power draw and heat if these systems suddenly kick the GPU's into gear. Obviously, assuming BOINC is updated on these systems too.

Yes this is correct the school does not operate 365 days a year. I think I have the numbers hashed out correctly.
I am now actually trying to write a script that will change the way the WCG client behaves when it is connected to the schools network, and when it is connected to any other network.
For example 40% usage while on campus and plugged in, versus 70% usage while anywhere else and plugged in. I did not see a way to have the clients "switch" profiles using the profile manager
I am also getting ready to deploy the WCG program to the clients and have used orca to edit the msi installer and create a transform, but the clients are not automatically connecting to my account, any suggestions?

Saint Marys School ,

You might want to have a real hard look at the ThreadMasterGUI tool for Windows [link in Start Here FAQ's]. The BOINC throttle is poorly: at 40% it runs 2 seconds at 100%, then pauses 3. At 70% it runs 7 seconds, then pauses 3. It does not know the fractional second concept, which TMG does... set a default ceiling of 70 and 40%, then set the OS scheduler to execute.... runs smooth at microsecond level. BOINC's throttle design was to allow sufficient cool down so laptops wont overheat but had implementation constraints as it has to work on all platforms it runs [1 code base]. I run TMG not so much to cool down, but merely to have a smooth crunch without the hacksaw CPU load.

See you're more than proficient to mass roll-out without WCG support, but still like you to contact WCG as large fleets can have impact on the supply and results, particular if something goes belly up, in style of course. They'd be able to ''tailor'' for your specific needs, even set up sub-team type contribution as a partner... involve the students, spread the word.

--//--

Edit: If at all, for laptops and to stop any fan revving, should that occur, 50% is best solution. 1 second alternations, usually never lets the fans gear every 2nd second. Stops fan wear, a mechanical piece.

If I had it right and supposing that we can consider a 2/3 active period that makes the total cost around 1'752 US$/year or 33.7 US$/week. I may have missed something but I am suprised that such an amount could be considered an "electrical bill shock".

Nevertheless if it is the case, then there is an excellent way to turn this into a participative effort by your students and ask them to contribute 1 US$ each, a symbolic but very positive effort.

Ok Well so far so good. I have a total of 30 devices joined now.
Have several students running the software and no issues so far.
I installed on a Friday and told the kids to make sure they left their computer plugged in and turned on........only one child actually did. So we will have to do an education session about the software and how it works.
Our marketing/PR people are working on a partner statement so we can become a partner.
Our server environment is almost completely virtualized, so we are running Vmware vsphere on 3 physical servers that host all of our virtual servers. With that being said only 2 of the servers are used at any one time, the 3rd server is a failover server. Since it is not utilized unless in a failover situation, I created some Ubuntu Virtual servers, and they are now running for the grid.
Since the server IS my failover server and is necessary for keeping my production machines running, I had to create some tasks that would handle the failover event in vsphere. Vsphere will try to gracefully shut down the grid computing servers, but if they are unable to shut down within 30 seconds it will shut them completely off so it can start migrating production servers.
I edited my UPS software to also shutdown the grid servers if running on battery backup.
I also had to add a new temperature sensor to monitor the server room temperature.
Back to the laptops, I intend to increase the total installations to 75 by tomorrow, and we will see how they run over the Christmas break, and we are preparing to to a push to get every kid here running BOINC/WCG to get at least one friend to install and run WCG.