[Opensim-dev] object memory footprint for ODE

[Nebadon Izumi]

just clarification i made error below and said our "stats" are server side,
it should read our physics are server side.

sorry about confusion there.

On Fri, Oct 31, 2008 at 8:14 AM, Nebadon Izumi <nebadon2025 at gmail.com>wrote:

> its not that simple, it depends on the complexity of your meshing, what
> shapes you chose etc.. a box wont use the same amount of memory as a torus
> thats cut twisted and hollowed out.   Also sculpty prims tend to use a bit
> more memory than standard prims,  its almost impossible to predict what
> memory it will use at this point, but chances are with 1gb ram you will be
> able to run 1 sim with 15k prims. all depends, if your running linux likely
> 1gb is not even enough ram to run 1 full 15k prim sim.. on the other hand in
> windows you could likely run 1 sim with 15k prims and have some memory to
> spare.. there are just to many variables in the equation though to predict
> accuratly.   and again, our stats are server side, cuda wont do you any
> good, its likely opensim will never be able to use video card accelerated
> physics, its just not possible short of re-writting the linden viewer from
> scratch and redoing opensim 100% also.
>
> Neb
>
>
> On Fri, Oct 31, 2008 at 8:09 AM, Eugen Leitl <eugen at leitl.org> wrote:
>
>> On Fri, Oct 31, 2008 at 08:02:03AM -0700, Nebadon Izumi wrote:
>> >
>> >    physics are done server side on opensim and do not ever use video
>> card
>>
>> To clarify, I'm looking at hacking ODE to support CUDA acceleration.
>>
>> >    memory, so the answer to the video card memory is it use 0mb of ram.
>> >    OpenSIM and Secondlife all use server side phyiscs.
>>
>> Assume 1 GByte available for ODE. How high scene complexity can
>> I expect?
>>
>> --
>> Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
>> ______________________________________________________________
>> ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
>> 8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
>> _______________________________________________
>> Opensim-dev mailing list
>> Opensim-dev at lists.berlios.de
>> https://lists.berlios.de/mailman/listinfo/opensim-dev
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://opensimulator.org/pipermail/opensim-dev/attachments/20081031/9c12cc43/attachment-0001.html>