[Bf-cycles] Tiles sizes - impact on RAM and speed

[Mickaël Bourgeoisat]

Hi, if it's one pixel per core for best gpu time rendering, the tile 
size must be set also to the same pixel ratio as the output frame ratio 
divided by the number of core (and we must try to obtain a %16 XY), is 
that right ???
Is that rule the same for kepler and fermi ?
So there is no "generic" best tile size for every one, it's adaptative 
to the CG ...
But for cpu ? With only height or four cores ? which are really 
different from CG ?
Thanks a lot for your answers :)
Myke

Le 11/02/2013 00:40, Cal McGaugh a écrit :
> I still don't understand why tiles are better than progressive renders 
> (?)
> regardless of CPU or GPU.
> Much less which tile size/resolution settings are best for any given
> scene.
> I don't do complicated scenes.....mostly fluid sims. But like to animate
> using Cycles, so getting the fastest render time is very important to me.
> It would be great if Blender Cycles "sensed" the optimal settings 
> depending
> on the resolution and other variables were chosen. Possible?
> BTW, thanks Ole for the FAQs......will study them.
> And to Brecht for all your work on Blender.
> Cal
>
> On February 10, 2013 at 2:42 PM Ole Jakob Skjelten <olesk at pvv.org> wrote:
>> Oh sorry guys, one last thing. As I mentioned initially, these 
>> questions were for an article in a Cycles GPU Hardware FAQ I've been 
>> setting up trying to preempt the most common questions that gets 
>> asked repeatedly at Blenderartist regarding all things hardware 
>> related to GPU rendering.
>> If anyone wants to copy & paste or link parts or all of it to the 
>> official Blender wiki you are of course more than welcome to. As 
>> previously mentioned, you can find it at 
>> http://www.systemagnostic.com/faqs/
>> OJ
>>
>> -- 
>> Ole Jakob Skjelten
>> olesk at pvv.org <mailto:skjelten at wegelin.ch>
>> +41 79 832 2976
>>
>>
>>
>> On 10. feb. 2013, at 18:45, Brecht Van Lommel < 
>> brechtvanlommel at pandora.be <mailto:brechtvanlommel at pandora.be>> wrote:
>>
>>> On Sun, Feb 10, 2013 at 5:52 PM, Ole Jakob Skjelten < olesk at pvv.org 
>>> <mailto:olesk at pvv.org>> wrote:
>>>> Thanks a lot for that Brecht! My understanding is that *reducing* 
>>>> tile sizes
>>>> only has a positive effect with regards to reducing the size of the 
>>>> output
>>>> buffer, which is in most cases is a rather small part of the total 
>>>> RAM eaten
>>>> by Cycles, and thus, unless you you're in a desperate
>>>> I-only-need-10MB-more-RAM-to-render-on-my-GPU scenario, increasing 
>>>> the tile
>>>> size for speed benefits is the more likely scenario. Would you agree? 
>>>
>>> Yes, I agree. Maybe if you have a ton of render passes, but otherwise
>>> you can quite safely set it quite large.
>>>
>>>> It's also my understanding that, using CUDA, keeping your tiles to 
>>>> multiples
>>>> of 16 pixels helps speed (or rather prevent wasting GPU cycles)? And
>>>> finally, does it help to operate with tiles that are multiples of 
>>>> the full
>>>> render resolution, or should one use same values for x and y direction
>>>> (square tiles) instead? 
>>>
>>> Multiples of 16 might indeed help, I'm not quite sure if this true in
>>> general though and if this perhaps depends on the particular GPU.
>>> Multiples of the render resolution will help if you use big tiles,
>>> since otherwise you can get some smaller tiles to fill up the full
>>> render.
>>>
>>> Brecht.
>>> _______________________________________________
>>> Bf-cycles mailing list
>>> Bf-cycles at blender.org <mailto:Bf-cycles at blender.org>
>>> http://lists.blender.org/mailman/listinfo/bf-cycles 
>
>  _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> http://lists.blender.org/mailman/listinfo/bf-cycles
>
>
> _______________________________________________
> Bf-cycles mailing list
> Bf-cycles at blender.org
> http://lists.blender.org/mailman/listinfo/bf-cycles

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.blender.org/pipermail/bf-cycles/attachments/20130211/f5470064/attachment.htm