Is there a way for you to set a limit on what projects you receive?

For example, set the project so that you only receive projects totaling 2000 hours of computing time or less?

Thanks.

No.
There is (currently), only 1 type of model, and that is 160 years long.
Other types may show up later in the year, after LOTS of testing.

On about 4th of December each model year, some climate data is returned with the trickle.
Every 10 model years, a larger amount of data is also returned as a zip file.
Every 40 model years, a \'restart dump\' is returned.
Also, the first half of the modelling is \'hindcast\', and the second half is \'forecast\'.

So the 160 year models can be thought of as:
160 1 year models
16 10 year models
4 40 year models
2 80 year models, 1 hindcast, 1 forecast

It\'s my understanding that all the climate change projects are the same size - the length of time they take depends on the speed of your computer and how much it is switched on.

---

Visit the Scotland team

It\'s my understanding that all the climate change projects are the same size - the length of time they take depends on the speed of your computer and how much it is switched on.

I am familiar with this (that all models span the same amount of \"years\", but I am in the dark as to why the projects I have received have had estimated \"To Completion\" times of anywhere from 1000 to 5000 hours. Is it because some models are simply more complicated than others (even thought they model the same lenght of time)?

Another issue that I would like to have answered, is whether your project will be terminated and not evauluated if it takes your computer more time than estimated, and therefore it is sent back to the server after its \"Report Deadline\".

Thanks.

The estimates are based on Boinc\'s benchmarks, and Boinc\'s benchmarks are frankly useless.

You can get a much better estimate by calculating it yourself - look at the \'% completed\' figure, and the \'CPU hours so far\'. Divide CPU hours by the amount completed.

For example: if % completed is 1%, and CPU hours is 15, then 15/0.01 = 1,500 hours.

The report deadline is ignored by this project. You can take as long as you want. furthermore, the models climate is uploaded to the servers at intervals, so it doesn\'t matter if you can\'t reach all the way to the end of the model.

Take a look at the READMEs I linked to in the other thread, it\'ll give you some additional background information.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

When you first start a model, it takes BOINC a while to calculate an estimated completion time for your particular computer - it will probably take several trickles from you before BOINC works out seconds per timestep and produces anything like a good estimate.

The deadlines for completion are a standard feature of BOINC projects, most of which are very much shorter than the climate models. Although deadlines continue to be shown for the climate models, they are of no significance and can safely be ignored (so long as you continue to trickle, so that the server knows you are still crunching).

Good luck!
[Edit: oh, Mike beat me to it!]

---

Visit the Scotland team

Along with what the others have said, there are a several other things in the calculation.

How many projects you are running.
The Result Duration Correction Factor.
How often BOINC thinks your computer is on, and how many hour a day it is allowed to run projects.
Benchmarks
Version of BOINC (yes this makes a difference)

It\'s not just a simple answer to give you an estimate.

---

Along with what the others have said, there are a several other things in the calculation.

How many projects you are running.
The Result Duration Correction Factor.
How often BOINC thinks your computer is on, and how many hour a day it is allowed to run projects.
Benchmarks
Version of BOINC (yes this makes a difference)

It\'s not just a simple answer to give you an estimate.

Why would a different version of the BOINC Manager help speed up(or perhaps slow down) the rate at which a project runs?
Also, doesn\'t the BOINC Manager give special priority to a projects if it determines that running the project at it\'s alotted percentage won\'t be enough for it to finish on deadline?

Why would a different version of the BOINC Manager help speed up(or perhaps slow down) the rate at which a project runs?
Also, doesn\'t the BOINC Manager give special priority to a projects if it determines that running the project at it\'s alotted percentage won\'t be enough for it to finish on deadline?

Because the different versions take different parts into the calculation of time. Older ones may not look at RDF, or they benchmark differently, or...

Each one has their own personality. They have been adjusted, changed, etc. So sometime the estimates are totally different from version to version.

---