 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Dear all,
This is probably not directly relevant to the vast majority of people
here, but I just thought I'd share what I've been using pov for. Even if
it isn't directly relevant to you, you can see some pretty-ish pictures :)
My work involves integrated circuit design. You can get some pretty
impressive pictures when you print an entire chip design, but I thought
it'd be even better to be able to make 3D pictures.
So, I've written a program to convert the chip layout data from the
standard gds2 format to a (usually large!) pov scene file.
You probably understand why it isn't directly relevant to most people
here - only a small number of people will be able to generate the
necessary data.
Anyway, if anybody does want it, or else wants to see some of the
pictures I've generated, take a look at:
http://www.atchoo.org/gds2pov/
gds2pov is available for Windows, Linux (x86) and Solaris (Sparc).
I hope you like it :)
Cheers,
Roger
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey Roger thats real cool.
I liked that Octagonal image you have on the site. Any chance you could
post the code for it. I think it would be real fun to try re-texturing and
relighting it.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
news:41189ace$1@news.povray.org...
> You probably understand why it isn't directly relevant to most people
> here - only a small number of people will be able to generate the
> necessary data.
Actually these would make fantastic landscapes with some fancy texturing and
lighting. How is it possible to produce GDS files (with no intent whatsoever
at chip layout...) ?
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Cool! A couple of years ago for the world within worlds IRTC round I hit
upon the idea of rendering one of the chips we'd designed which was inside
the box I used to do the render. Image was a cloudy semi transparent sphere
which contained a rendering of the entire chip inside the sphere floating
over the chip itself in large. Kind of the sphere over the checker plane
twisted on in on itself. My management loved it and approved entering the
contest, but our legal department nixed it and the image as well as several
other derivative images has never seen the light of day - I did make a
couple of personal prints which hang in my apartment.
I converted from a format similar to gds though to do the whole chip - which
was very large - I transitioned to material maps once some distance from the
look at position. It looks like some images on your site are real chip data
and others possibly not or at least a layout style/process unfamiliar to me.
If it is real chip data, take care that you have been granted legal
permission to use the mask data and release images of it. Companies invest
10s of millions in a set of mask data and they can be pretty protective of
it!
"Roger Light" <rog### [at] example net> wrote in message
news:41189ace$1@news.povray.org... gds2pov is available for Windows, Linux
(x86) and Solaris (Sparc).
> Roger
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Gilles Tran wrote:
> news:41189ace$1@news.povray.org...
>
>>You probably understand why it isn't directly relevant to most people
>>here - only a small number of people will be able to generate the
>>necessary data.
>
>
> Actually these would make fantastic landscapes with some fancy texturing and
> lighting. How is it possible to produce GDS files (with no intent whatsoever
> at chip layout...) ?
Well to be honest, if you're just wanting to make nice landscapes you'd
probably be better off just doing it in pov!
All I have there are essentially lots of prisms and lots of boxes
(although the boxes are actually implemented as mesh2 objects).
The different colours indicate different layers in the chip, different
metal conductor layers and their interconnects for example, with the
height and thickness being fixed for any given layer.
Having said that, you could try Magic (
http://bach.ece.jhu.edu/~tim/programs/magic/ ) or Electric (
http://www.gnu.org/software/electric/ ) if you are on Unix, or Lasi (
http://members.aol.com/lasicad/ ) if you are on Windows.
Cheers,
Roger
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
William Pokorny wrote:
> Cool! A couple of years ago for the world within worlds IRTC round I hit
> upon the idea of rendering one of the chips we'd designed which was inside
> the box I used to do the render. Image was a cloudy semi transparent sphere
> which contained a rendering of the entire chip inside the sphere floating
> over the chip itself in large. Kind of the sphere over the checker plane
> twisted on in on itself. My management loved it and approved entering the
> contest, but our legal department nixed it and the image as well as several
> other derivative images has never seen the light of day - I did make a
> couple of personal prints which hang in my apartment.
Wow, that sounds really good. Shame you couldn't release it!
> If it is real chip data, take care that you have been granted legal
> permission to use the mask data and release images of it. Companies invest
> 10s of millions in a set of mask data and they can be pretty protective of
> it!
As you say, most of the pictures aren't "real" data. The ones that are
real data are not particularly sensitive plus you can't really see that
much. Distributing the scene code would be another thing altogether of
course.
The work you see is my own, with the exception of the very complicated
looking scene. I am in research at a university so we don't really have
that many trade secrets.
Cheers,
Roger
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hi Josh,
> I liked that Octagonal image you have on the site. Any chance you could
> post the code for it. I think it would be real fun to try re-texturing and
> relighting it.
Thanks :) It is interesting that I have had more replies to my post here
than I did at the cad newsgroup I posted to.
I can't really post code for any of the scenes there. They do contain
the actual layout information so, yeah, I don't want to give that out!
Sorry,
Roger
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"William Pokorny" <pok### [at] attglobalnet> wrote:
>
> My management loved it and approved entering the contest,
> but our legal department nixed it and the image as well
> as several other derivative images has never seen the
> light of day - I did make a couple of personal prints
> which hang in my apartment.
>
A reasonable expectation for pre-GA'ed hardware. Would you have been
allowed to do so using really old masks-- like something from 0.25 um
generation? Using "kerf macros" instead of customer "product areas" ?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Not sure. Many of the concerns raised had to do with simply losing control
of how the image would be used if submitted to the competition. There was an
entire list of legal concerns with the irtc submission agreement. In play
too is they way big organizations work. Lawyers charged with keeping the
company out of legal trouble are not likely to take any risk on something as
relatively unimportant as the submission of an image to the IRTC
competition. It is important to note too the legal department didn't just
say no. They evaluated the situation, the submission agreement and came
back with a detailed response as to why they would not grant permission. I
could understand many of their concerns.
"gregjohn" <pte### [at] yahoocom> wrote in message > >
>
> A reasonable expectation for pre-GA'ed hardware. Would you have been
> allowed to do so using really old masks-- like something from 0.25 um
> generation? Using "kerf macros" instead of customer "product areas" ?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
