 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3C1F2005.5687CA96@free.fr> , Adrien Beau <adr### [at] free fr>
wrote:
> I don't know what this means, but I do know that DDR is better
> at working on huge amounts of contiguous data, such as gigantic
> meshes, enormous image maps, etc.
Well, POV-Ray will never access them in a continuous fashion but randomly so
you gain little with DDR-DRAM in this case :-(
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3c1f55ec@news.povray.org> , Warp <war### [at] tagpovrayorg> wrote:
> On the other hand, when there's a cache miss, significant amount of time is
> spent copying a portion of the RAM to the cache. If the RAM is faster,
> wouldn't this copying be faster as well?
Yes and no. It all depends very much on the cache and processor design.
Usually cache misses will be hidden due to pre-fetching. In particular x86
processors have extraordinarily long pipelines with ten or more stages even
for simple instructions. So those can hide a lot of latency in memory if
the processor reads early and there are numerous other methods to cover for
slow memory.
In addition statistics show that most thinks at a particular time are done
locally on a small scale. I.e. a MPEG/JPEG discrete cosine transform will
take blocks of 8*8, so you have 64 fairly local accesses again. So even in
extreme data processing applications like these there is some locality The
same is true for POV-Ray on the small scale, i.e. when walking a linked
structure.
While this may seem like a contradiction, a good cache architecture like the
ones used in processors today will hide almost all of these different yet
similar access types and main memory access plays only a very minor role as
long as the time it takes to process one piece of information from memory
still takes longer than it takes to get it from memory.
Well, the whole situation is very complex with the cache and processor
designs today because of out-of-order and delayed bus transfers or even
packet based bus systems and hard to explain without writing a book about
it, so I will stop here...
Thorsten
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thanks for all of the dialogue. I have also been asking others and the
consensus seems to be that more processing power will bring the biggest
bang for the buck although there could be some improvement with faster
memory. I'll probably go with an 1800+ and overclock it a bit and save the
money for DDR and go with sdram 133.
Now my most exciting discovery ... in the midst of this discussion (and
others that I have been having) I stumbled on a web page that hosts a
benchmark and a recompiled version of POVRay 3.1g that is optimized for the
PIII and Athlon, it was done as an experiment. I took a low resolution
render of one of my images (160X120 no AA) and rendered it. Standard POVRay
3.1g took 15 min and 47 sec. With the code optimized it took 4 min and 0
sec. I was astounded. Combining this with a faster processor, I think there
will be a great improvement.
Now, because of the complexity of the image, most of the time saved with the
optimized POV engine was in the parsing. I'll let you know an approximated
time savings after I render a few of the pictures at full resolution where
most of the time saved will be in actual rendering.
For those interested in the site, it is located at:
http://students.washington.edu/sschmitt/pov/
Kirk
"Kirk Hubbell" <kir### [at] yahoocom> wrote in message
news:3c1ea50e@news.povray.org...
> I am building a new PC for rendering images with POVRay. I am working on a
> project that will require a lot of images. I have already done the work
but
> I still need to render. Due to the complexity of the images each image
takes
> over a day to render, so speed is important.
>
> The problem is my budget.
>
> Question, for the Windows version of POVRay has anybody done a benchmark
> comparison of how fast images render with the Althon XP processor with 133
> sdram versus DRR 2100?
>
> I know the DDR is faster but I have heard that streaming applications (not
> computative intensive ones) are the big winners.
>
> I guess that I really want to know if it would be worth the money to
upgrade
> to the DDR memory if my other choice were to upgrade to a 1800 or 1900
> Althon (assuming that I could not afford to do both).
>
> I have looked at some POVRay bench mark sites but they do not distinguish
> sdram/DDR in an apples to apples comparison.
>
> Thanks for any help.
>
> Kirk
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
One last note:
After doing several test renders with the optimized POVRay3.1g from the
website below it looks like parsing improves by about 400% and rendering
improves by 20% to 25%, I didn't do a detailed measurement.
Kirk
"Kirk Hubbell" <kir### [at] yahoocom> wrote in message
news:3c1ff1e7@news.povray.org...
> Thanks for all of the dialogue. I have also been asking others and the
> consensus seems to be that more processing power will bring the biggest
> bang for the buck although there could be some improvement with faster
> memory. I'll probably go with an 1800+ and overclock it a bit and save the
> money for DDR and go with sdram 133.
>
> Now my most exciting discovery ... in the midst of this discussion (and
> others that I have been having) I stumbled on a web page that hosts a
> benchmark and a recompiled version of POVRay 3.1g that is optimized for
the
> PIII and Athlon, it was done as an experiment. I took a low resolution
> render of one of my images (160X120 no AA) and rendered it. Standard
POVRay
> 3.1g took 15 min and 47 sec. With the code optimized it took 4 min and 0
> sec. I was astounded. Combining this with a faster processor, I think
there
> will be a great improvement.
>
> Now, because of the complexity of the image, most of the time saved with
the
> optimized POV engine was in the parsing. I'll let you know an approximated
> time savings after I render a few of the pictures at full resolution where
> most of the time saved will be in actual rendering.
>
> For those interested in the site, it is located at:
>
> http://students.washington.edu/sschmitt/pov/
>
> Kirk
>
> "Kirk Hubbell" <kir### [at] yahoocom> wrote in message
> news:3c1ea50e@news.povray.org...
> > I am building a new PC for rendering images with POVRay. I am working on
a
> > project that will require a lot of images. I have already done the work
> but
> > I still need to render. Due to the complexity of the images each image
> takes
> > over a day to render, so speed is important.
> >
> > The problem is my budget.
> >
> > Question, for the Windows version of POVRay has anybody done a benchmark
> > comparison of how fast images render with the Althon XP processor with
133
> > sdram versus DRR 2100?
> >
> > I know the DDR is faster but I have heard that streaming applications
(not
> > computative intensive ones) are the big winners.
> >
> > I guess that I really want to know if it would be worth the money to
> upgrade
> > to the DDR memory if my other choice were to upgrade to a 1800 or 1900
> > Althon (assuming that I could not afford to do both).
> >
> > I have looked at some POVRay bench mark sites but they do not
distinguish
> > sdram/DDR in an apples to apples comparison.
> >
> > Thanks for any help.
> >
> > Kirk
> >
> >
>
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Please do not quote the whole of a message when replying.
Most if not all of us have the previous message.
"Kirk Hubbell" <kir### [at] yahoocom> wrote in message
news:3c1ff1e7@news.povray.org...
> Thanks for all of the dialogue. I have also been asking others and the
> consensus seems to be that more processing power will bring the biggest
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3c215c74@news.povray.org> , "Kirk Hubbell" <kir### [at] yahoocom>
wrote:
> After doing several test renders with the optimized POVRay3.1g from the
> website below it looks like parsing improves by about 400% and rendering
> improves by 20% to 25%, I didn't do a detailed measurement.
But this is nothing new. Of course the latest Intel compiler code is better
than the (three year old) Visual C++ or Watcom compiler code. Just take a
look at the POV-Ray 3.5 beta, which uses the Intel compiler.
Thorsten
____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
