 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Micha Riser wrote:
>
> As I understood it, 3DNow supports single precision floating point only.
> But POV (and ray-tracing in general) needs double precision to get good
> pictures (there is a thread in p.unofficial.patches about why this is
> so). Therefore 3DNow is useless for POV.
>
> - Micha
Specifically the threads:
3D!Now & ISSE
double/single precision
BTW, wouldn't that be a good addition to the VFAQ?
Christoph
--
Christoph Hormann <chr### [at] gmx de>
IsoWood include, radiosity tutorial, TransSkin and other
things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tom Melly wrote:
>
> Question: If 3DNow speeds up float calcs, why is it only of benefit to
> graphics and not all apps. using floating point calcs.?
WinAmp uses 3DNow to decode mp3s.
______________________________________________________________________
Kari Kivisalo http://www.kivisalo.net
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Kari Kivisalo" <kar### [at] kivisalonet> wrote in message
news:3A717C79.10257149@kivisalo.net...
> Tom Melly wrote:
> >
> > Question: If 3DNow speeds up float calcs, why is it only of benefit to
> > graphics and not all apps. using floating point calcs.?
>
> WinAmp uses 3DNow to decode mp3s.
>
Yes - I gathered that it applied to multi-media (hate that term) - this is
partly what confused me. MM as a generic set of apps is hardly likely to
benefit from something unless that something can also benefit non-MM apps.
The answer would appear to be the 3dnow only works on single-point.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Just have to push to get AMD to beef up 3DNow with double precision, I
guess. Could call it Super Extended 3DNow! :)
And yes, it would be a good addition to the FAQ.
"Christoph Hormann" <chr### [at] gmxde> wrote in message
news:3A717642.3C96091C@gmx.de...
>
>
> Micha Riser wrote:
> >
> > As I understood it, 3DNow supports single precision floating point only.
> > But POV (and ray-tracing in general) needs double precision to get good
> > pictures (there is a thread in p.unofficial.patches about why this is
> > so). Therefore 3DNow is useless for POV.
> >
> > - Micha
>
> Specifically the threads:
>
> 3D!Now & ISSE
> double/single precision
>
> BTW, wouldn't that be a good addition to the VFAQ?
>
> Christoph
>
> --
> Christoph Hormann <chr### [at] gmxde>
> IsoWood include, radiosity tutorial, TransSkin and other
> things on: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tom Melly wrote:
> "Kari Kivisalo" <kar### [at] kivisalonet> wrote in message
> news:3A717C79.10257149@kivisalo.net...
> > Tom Melly wrote:
> > >
> > > Question: If 3DNow speeds up float calcs, why is it only of benefit to
> > > graphics and not all apps. using floating point calcs.?
> >
> > WinAmp uses 3DNow to decode mp3s.
> >
>
> Yes - I gathered that it applied to multi-media (hate that term) - this is
> partly what confused me. MM as a generic set of apps is hardly likely to
> benefit from something unless that something can also benefit non-MM apps.
> The answer would appear to be the 3dnow only works on single-point.
My understanding of 3dnow, and intel's counterpart Katimi new instructions, is
that it merely allows you to do the same instruction on several pieces of data
simultaneously. Basically it saves clock cycles by not having to repeat the
same instruction. Example multiplying normally you would input something like
this into the processor: data, data, multiply, data, data, multiply, data,
data, multiply... But with (single instruction multiple data) it would use a
single command to to input all the multiply's. Well this is my understanding
of it. I have never read the actual list of commands an intel processor
recognizes.
Dan Johnson
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
http://www.students.tut.fi/~warp/povVFAQ/miscVFAQ.html#3dnow
--
char*i="b[7FK@`3NB6>B:b3O6>:B:b3O6><`3:;8:6f733:>::b?7B>:>^B>C73;S1";
main(_,c,m){for(m=32;c=*i++-49;c&m?puts(""):m)for(_=(
c/4)&7;putchar(m),_--?m:(_=(1<<(c&3))-1,(m^=3)&3););} /*- Warp -*/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On 2001-01-26 10:31, Michael Brown <emb### [at] i4freeDELETETHISconz> wrote:
>Just wondering if Povray supports 3DNow (plain vanilla K6-2 and/or extended
>Athlon), and if not if it plans to support it in the future.
Look in povray.unofficial.patches. Daniel Jungmann has converted the
color handling (which is single precision) to use 3D!Now and reports a
speedup of at least 50%.
hp
--
_ | Peter J. Holzer | All Linux applications run on Solaris,
|_|_) | Sysadmin WSR | which is our implementation of Linux.
| | | hjp### [at] wsracat |
__/ | http://www.hjp.at/ | -- Scott McNealy, Dec. 2000
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I wish he were more precise about how his optimizations speed things up.
What is it exactly that has changed? Are *all* renders now just magically
50% faster, or what?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Tony[B]" wrote:
> Are *all* renders now just magically 50% faster, or what?
No, not likely. Intersection testing will remain the same as well
as features such as media and radiosity. Complex textures and color
combinations will benefit as well as a possible speed up of multiple
colored light sources. Without having had a chance to test the
optimizations it is difficult to predict how much faster the "average"
scene file will be but you don't get something for nothing. For
example if you don't have a processor with 3DNow support you are out
of luck :)
--
Ken Tyler
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> For example if you don't have a processor
> with 3DNow support you are out of luck :)
Should there be significant improvement via his modifications to the source,
I would not hesistate to obtain one. :)
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
