Subject says it all. See http://www.povray.org/beta/ for the details.

-- Chris

Post a reply to this message

Feel like I'm left behind in the dust with my current, and what I consider
to still be new, PC; having only the Intel Pentium IV with Hyperthreading
and my WinXP being only 32bit, too, I keep thinking I'm always making bad
computer buys.
:-)
I'm guessing this new and improved POV-Ray will be okay on a 64bit Intel
processor like the Nocona, or whatever it will be called...? But why bother,
since AMD seems to be staying ahead on that processing power stuff.
Although, having checked up on pricing they don't seem to be very different
in cost and Intel is boasting a lot of improvements.

Regardless, I'm bound to be a few steps behind the top-end system users out
there no matter what I try to do. Heh-heh-heh.

Bob H.

Post a reply to this message

Hughes, B. wrote:
> Feel like I'm left behind in the dust with my current, and what I consider
> to still be new, PC; having only the Intel Pentium IV with Hyperthreading
> and my WinXP being only 32bit, too, I keep thinking I'm always making bad
> computer buys.
> :-)
> I'm guessing this new and improved POV-Ray will be okay on a 64bit Intel
> processor like the Nocona, or whatever it will be called...? But why bother,
> since AMD seems to be staying ahead on that processing power stuff.
> Although, having checked up on pricing they don't seem to be very different
> in cost and Intel is boasting a lot of improvements.

This 64bit version is mainly developed for the AMD 64bit processors 
(Athlon 64/Opteron).  Of course it should also run on whatever Intel 
develops as compatible to the AMD64 extensions.  It won't run on Intel's 
Itanium systems but that's a different league anyway.

Since 64bit Windows versions are still in beta the whole 'Windows 
64-bit' concept is not that important for the broad market yet, the 
number of people running POV-Ray on 64bit Unix systems (including AMD64) 
is probably going to be higher than on Windows for quite some time.  If 
you want to buy a relatively cheap and reliable 64bit system for use 
with POV-Ray, AMD with Linux/FreeBSD is probably the way to go at the 
moment.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 06 Jul. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Chris Cason wrote:
> Subject says it all. See http://www.povray.org/beta/ for the details.
> 

BTW there just has been an article on slashdot on a comparison of AMD64 
systems on Linux/Windows:

http://slashdot.org/articles/04/07/12/117254.shtml

The tests include POV-Ray, but only with the 32bit version on Windows. 
The quite different numbers with the different Linux distributions 
indicate that they used the POV-Ray versions coming with the distributions:

http://anandtech.com/linux/showdoc.aspx?i=2114&p=4

The sense of this test is of course extremely questionable (what are 
they actually trying to compare there?).

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 06 Jul. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Christoph Hormann <chr### [at] gmxde> wrote:
> The sense of this test is of course extremely questionable (what are
> they actually trying to compare there?).
>
> Christoph

Exactly what they state they are testing: the readiness of certain operating
systems for 64bitness "out of the box".  That is, no recompiling, no
additional software to make things work.  You can, of course get much
better performance under Linux by recompiling software to suit your tastes,
but that wasn't the point of the article.

....Chambers

Post a reply to this message

What kind of performance changes/gains can we expect from the 64-bit
release?  Anyone running benchmarks yet?

Skip

Post a reply to this message

Chambers wrote:
> 
>>The sense of this test is of course extremely questionable (what are
>>they actually trying to compare there?).
>>
>>Christoph
> 
> 
> Exactly what they state they are testing: the readiness of certain operating
> systems for 64bitness "out of the box".

No, for that they would not need any benchmark results because they have 
nothing to do with the 'readiness'.  The only 'readiness' fact that you 
can indirectly deduce from the results is the availability of 64bit 
software on the different systems.

What you can conclude from the benchmark results they gave is: how the 
official 32bit version on 64bit Windows compares to the 32bit version on 
two specific 64bit Linux distributions - basing an OS/hardware decision 
on that would be just insane.  And you can see which of the two tested 
Linux distributions ships a better optimized 64bit POV-Ray which also 
does not have much to do with which of them is more 'ready'.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 06 Jul. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Skip Talbot <sta### [at] uiucedu> wrote:
> What kind of performance changes/gains can we expect from the 64-bit
> release?  Anyone running benchmarks yet?

  The AMD64 processor has additional features (eg. additional registers)
only available in 64-bit mode which the compiler can use to generate
faster code than when compiling to 32 bits. Thus it is expected that
the 64-bit version will run faster in an AMD64 than the 32-bit version
(how much faster depends on the compiler used).
  If you have an AMD64, it just doesn't make sense to use the 32-bit
version of POV-Ray because the 64-bit version should be faster.

  (And by the way, the 64-bit version being faster does not mean that
64-bitness in itself brings magically some speed benefit. The speedup
is not related to the 64-bitness of the processor, but to the fact that
in 64-bit mode it has additional features not available in 32-bit mode.)

-- 
#macro N(D)#if(D>99)cylinder{M()#local D=div(D,104);M().5,2pigment{rgb M()}}
N(D)#end#end#macro M()<mod(D,13)-6mod(div(D,13)8)-3,10>#end blob{
N(11117333955)N(4254934330)N(3900569407)N(7382340)N(3358)N(970)}//  - Warp -

Post a reply to this message

"Warp" <war### [at] tagpovrayorg> wrote in message news:40f39b77@news.povray.org...
>   (And by the way, the 64-bit version being faster does not mean that
> 64-bitness in itself brings magically some speed benefit. The speedup
> is not related to the 64-bitness of the processor, but to the fact that
> in 64-bit mode it has additional features not available in 32-bit mode.)

Precisely. I saw a speedup of about 10%, and there'll be more to come as
the compiler I am using is still in beta.

Plus 64-bit mode also allows you to render insane scenes ;-). (I've rendered
one here which used 3.5 gigabytes of RAM).

-- Chris

Post a reply to this message

Chris Cason <new### [at] deletethispovrayorg> wrote:
> Plus 64-bit mode also allows you to render insane scenes ;-). (I've rendered
> one here which used 3.5 gigabytes of RAM).

  Theoretically a 32-bit version could also do that, but it's very
possible that some OSs limit the maximum allocatable size to 2 GB.

  (Well, if we go deeper into theory, AFAIK you can use up to 64 GB of
memory with a Pentium or newer 32-bit processor, but the application
would need to be specifically coded to do that. A standard C++ program
can't do it. Naturally the OS has to support that as well (unless the
program bypasses the OS, if it's possible).)

  But I suppose that with a 64-bit binary the theoretical limits are
so immensely high that we don't need to worry about them in the next
several decades (hardware other than the CPU is and will be the limiting
factor for long time). Thus allocating 3.5 or 35 or even 350 gigabytes
of memory is not any problem in a 64-bit system (as long as the hardware
is up to those sizes)... :)

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message