Just an experiment with a macro that stitches
together POV-Ray's built-in bicubic patches.

Parsing: 1m 42s
Tracing: 9m 05s

@ 2GHz Athlon

Peak memory used: 903 MB


Tor Olav

Post a reply to this message

Attachments:
Download 'Patch03_06.jpg' (66 KB)

Preview of image 'Patch03_06.jpg'

Tor Olav Kristensen <tor_olav_kCURLYAhotmail.com> wrote in 
news:3fef91a8@news.povray.org:

> 
> Just an experiment with a macro that stitches
> together POV-Ray's built-in bicubic patches.
> 
> Parsing: 1m 42s
> Tracing: 9m 05s
> 
> @ 2GHz Athlon
> 
> Peak memory used: 903 MB

I ran it again with the type for all the bicubic patches
set to 0 instead of 1. Now I got these results:

Parsing: 18s
Tracing: 23s

Peak memory used: 12 MB


There's no I'm not able to see any differences between
the two images. (I wasn't expecting any either.)


Tor Olav

Post a reply to this message

"Tor Olav Kristensen" <tor_olav_kCURLYAhotmail.com> wrote in message
news:3fef91a8@news.povray.org...
>
> Just an experiment with a macro that stitches
> together POV-Ray's built-in bicubic patches.
>

Looks like it should be full of eggs.

Eric

Post a reply to this message

Eric Freeman wrote:
> Looks like it should be full of eggs.

eerie, that's what i thought

Post a reply to this message

In article <3fef96d5@news.povray.org>,
 Tor Olav Kristensen <tor_olav_kCURLYAhotmail.com> wrote:

> > Parsing: 1m 42s
> > Tracing: 9m 05s
> > 
> > @ 2GHz Athlon
> > 
> > Peak memory used: 903 MB
> 
> I ran it again with the type for all the bicubic patches
> set to 0 instead of 1. Now I got these results:
> 
> Parsing: 18s
> Tracing: 23s
> 
> Peak memory used: 12 MB

Interesting...how much memory does your system have? Swapping could 
account for a great deal of the slowness of type 1.

-- 
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: <chr### [at] tagpovrayorg>
http://tag.povray.org/

Post a reply to this message

Christopher James Huff <cja### [at] earthlinknet> wrote in news:cjameshuff-
735### [at] netplexaussieorg:

> In article <3fef96d5@news.povray.org>,
>  Tor Olav Kristensen <tor_olav_kCURLYAhotmail.com> wrote:
> 
>> > Parsing: 1m 42s
>> > Tracing: 9m 05s
>> > 
>> > @ 2GHz Athlon
>> > 
>> > Peak memory used: 903 MB
>> 
>> I ran it again with the type for all the bicubic patches
>> set to 0 instead of 1. Now I got these results:
>> 
>> Parsing: 18s
>> Tracing: 23s
>> 
>> Peak memory used: 12 MB
> 
> Interesting...how much memory does your system have? Swapping could 
> account for a great deal of the slowness of type 1.


I have (512-64)MB of RAM.
My PC is quite silent, so I didn't notice any serious disc-trashing.
But of course - you are right: It must have done some swapping.


Tor Olav

Post a reply to this message

"Eric Freeman" <whi### [at] yahoocom> wrote in 
news:3fefd4f2@news.povray.org:

> 
> "Tor Olav Kristensen" <tor_olav_kCURLYAhotmail.com> wrote in message
> news:3fef91a8@news.povray.org...
>>
>> Just an experiment with a macro that stitches
>> together POV-Ray's built-in bicubic patches.
>>
> 
> Looks like it should be full of eggs.

That gave me an idea for a scene. Thank you.


Tor Olav

Post a reply to this message

In article <3ff1174c$1@news.povray.org>,
 Tor Olav Kristensen <tor_olav_kCURLYAhotmail.com> wrote:

> > Interesting...how much memory does your system have? Swapping could 
> > account for a great deal of the slowness of type 1.
> 
> I have (512-64)MB of RAM.
> My PC is quite silent, so I didn't notice any serious disc-trashing.
> But of course - you are right: It must have done some swapping.

Yeah, it must have...512-64?
That's...<http://www.google.com/search?q=512-64>...448MB? Just a little 
under half as much as the scene was using...I'm guessing that made some 
difference. Want to run a test using about a third as many patches?

Hmm, systems with lots of memory bandwidth and space (like the PPC G5 or 
most recent Athlons) would probably do pretty well with type 1 patches...

-- 
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: <chr### [at] tagpovrayorg>
http://tag.povray.org/

Post a reply to this message

Christopher James Huff <cja### [at] earthlinknet> wrote in news:cjameshuff-
545### [at] netplexaussieorg:

> In article <3ff1174c$1@news.povray.org>,
>  Tor Olav Kristensen <tor_olav_kCURLYAhotmail.com> wrote:
> 
>> > Interesting...how much memory does your system have? Swapping could 
>> > account for a great deal of the slowness of type 1.
>> 
>> I have (512-64)MB of RAM.
>> My PC is quite silent, so I didn't notice any serious disc-trashing.
>> But of course - you are right: It must have done some swapping.
> 
> Yeah, it must have...512-64?
> That's...<http://www.google.com/search?q=512-64>...448MB?

You are right. 448 MB. That's because I have a motherboard with a
built in graphics "card", which is sharing the available memory
with my Athlon processor. I have allocated 64MB of my 512MB to the
graphics processor.


> Just a little 
> under half as much as the scene was using...I'm guessing that made some 
> difference. Want to run a test using about a third as many patches?

Ok. Now I have done so.

I reduced the number of patches to just above 5000.
Now the results are much better:

Parsing: 0m 09s
Tracing: 0m 07s

@ 2GHz Athlon

Peak memory used 315 MB.


> Hmm, systems with lots of memory bandwidth and space (like the PPC G5 or 
> most recent Athlons) would probably do pretty well with type 1 patches...

I think so too...


Tor Olav

Post a reply to this message