On Jul 22, 2009, at 10:35 AM, David Anderson wrote:

> Nicolás Alvarez wrote:
> ...
>> So it would distribute computing power across projects, which is  
>> what we want,
>> isn't it?
>
> Not exactly.  We want projects that are doing better or more  
> important science
> (as judged by volunteers) to get more computing power.
> The main criterion for choosing a project should be its scientific  
> merit,
> not how much credit it grants.

For the people who care most about the scientific merit of projects,  
credits are meaningless.  The cross-project parity problem doesn't  
apply to them.  They will be attracted to a project, and won't move  
regardless of the credit situation.

This discussion is about the *rest* of the crunchers, and how credits  
effect their behavior.

For this group of crunchers, they will concentrate in the top-paying  
projects, so long as we operate in the dreamland of cross-project  
parity.

Comparing only rank encourages equal distribution across all projects  
for this group of crunchers.


> This goal is called "credit neutrality".
>
> A second goal is that machines with special capabilities
> (e.g., large RAM, or ability to turn around jobs quickly)
> should have an incentive to participate in projects that
> need those capabilities.

If we didn't have to worry about cross-project parity, this could be  
done within a projects credit scheme.  Pay more credits for those  
resources you value.

> A proposal for achieving both of these goals is here:
> http://boinc.berkeley.edu/trac/wiki/CreditProposal
> (not likely to be implemented any time soon)

_______________________________________________
boinc_dev mailing list
[email protected]
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.

Reply via email to