yade-dev team mailing list archive

[Anton Gladky <gladky.anton@xxxxxxxxx>]

Hi, guys,

I think we should try to parallelize insertionsortcollider a little
bit and then measure productivity again. But not to brake the code, it
is better to do in separate branch, I think.


Anton




On Tue, Mar 1, 2011 at 9:31 PM, Sergei D. <sj2001@xxxxxxxxx> wrote:
>
>> Oh? ok. In the script you sent, it is not active (verletDist is negative
>> by default). By making it 0.3 radius, I can get down to 1 each 10
>> iterations.
>>
> But from documentation:
>
> "Negative value will trigger automatic computation, so that the real value
> will be |verletDist| × minimum spherical particle radius"
>
> So, Verlet is enabled in the script.
>
>> Valgrind always give interesting things, attached is some output for
>> InsertionCollider (with your script): 40% of collider time is spent in
>> the "find(id1,id2)" function, mostly because of shared_ptr manipulation.
>> A bit surprising...
>>
> Yes, surprising...
>
>
>
> --
> Best regards,
> Sergei D.
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~yade-dev
> Post to     : yade-dev@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~yade-dev
> More help   : https://help.launchpad.net/ListHelp
>