On Sat, 2005-05-07 at 10:02 +0200, Peter Karlsson wrote:
> On Fri, 6 May 2005, Daniel Phillips wrote:
> 
> > They have a risc processor on the board as well,  We won't have that.
> > But on the other hand, we have _way_ more gate equivalents.
> 
> Risc processor? Which project is this? The only one I know of is
> 'http://graphics.cs.uni-sb.de/~jofis/SaarCOR/DynRT/DynRT.html'
> which uses a fpga (don't know which brand)...

AFAIK, there are four FPGA-Raytracing projects out there.


The first one is the one you mentioned, the SaarCOR project that has
spawned the InTrace company. They are using a Virtex-II 6000-4 FPGA.

They can be found at http://www.saarcor.de/ or at
http://www.intrace.de/. The latest paper describing there work can be
found (as you said) at
http://graphics.cs.uni-sb.de/~jofis/SaarCOR/DynRT/DynRT.html.

They are no interested in Open Source or in cooperations. I have talked
to them at the CeBIT '04 and they were quiet secretive about their work.
If you look at their latest paper you may also find that a lot of
details are missing which therefore make it impossible to easily
reproduce their work.


The second one is one from Sanchez-Elez et al. You can find their '03
paper by googling for "Algorithm Optimizations and Mapping Schemes for
Interactive Ray Tracing on a Reconfigurable Architecture", which is the
title of their paper. Their seems to be no webpage describing their work
(or I have not found it). There is what seems to be a related
publication "Efficient mapping of hierarchical trees on coarse-grain
reconfigurable architectures" published in '04.


The third one is by Joshua Fender and Jonathan Rose. Their '03
publication's title is "A High-Speed Ray Tracing Engine Built on a
Field-Programmable System". The VHDL source code for that has been made
available freely here http://www.eecg.toronto.edu/~fender/ (under
"circuit benchmark sets"). The VHDL code is hardly documented and was
developed for a two-FPGA board called the "Transmogrifier-3". From my
short conversation with Mr. Fender I gather that this project is not
right now continued.


The forth project is my own ;). I'm doing a PhD in hardware raytracing.
I have no publications so far to show for it, but I'd be interested in a
cooperation and I'm also willing to open up at least parts of my code.

I'd be interested in one of the prototype cards without the OGP core and
without the linux driver - the bare card. I have all the necessary
development software available. However, time is pressing (as always),
so I'd want it ASAP - preferably yesterday. If I could get one soon, I
may be able to get University funding, otherwise I'd have to cough up
the money myself.


If you know of some other FPGA-Raytracing research/project, or if I made
a mistake in my description, feel free to inform/correct me. (Note that
the SaarCOR project has _lots_ of publications. Everything that has
Wald, Schmittler or Slusallek in the author list probably belongs in
there.)

Cheers,

-- Ulf

> Raytracing is one of my interest areas with the OGP card.
> 
> Best regards
> 
> Peter K

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to