Neither turn around time nor a high success rate is a resource usage.
Users that fail to meet the deadline take the risk of not getting credit -
that is the penalty. If there is a real world deadline, the project should
generate enough tasks to have a high probability of a quorum being met by
the real world deadline. The BOINC deadline for the project should be set
appropriately.
If GPU grid needs faster turn around time, they should tighten the deadline
a bit. Of course, they may lose some of the clients that way and have
lower throughput, but it is a tradeoff the project needs to make.
jm7
"Paul D. Buck"
<p.d.b...@comcast
.net> To
[email protected]
09/24/2009 01:02 cc
PM Martin <[email protected]>,
BOINC Developers Mailing List
<[email protected]>
Subject
Re: [boinc_dev] [boinc_alpha] Card
Gflops in BOINC 6.10
On Sep 24, 2009, at 9:22 AM, [email protected] wrote:
>> From your list, I would remove a few items.
>
> Turn Around Time would penalize dialup users and multi project
> users, and
> does not affect the final outcome at all. Of course, if work is
> returned
> after the deadline, it counts for 0, so there is a penalty already
> there.
> A low Success Rate has its own penalty - no credit for the time
> spent on
> the failures.
Actually it can, and, thus should be available for penalty and
reward. The current example is GPU Grid where the turn-around of work
is important to the project.
Secondly, it would also matter to your mythical 6 minute turn around
project ...
_______________________________________________
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.
