Jim,
I agree wholeheartedly and have gathered quite a lot of results myself and
posted them on our site:
http://www.tornadolabs.com/News/BenchJ3d_Results/benchj3d_results.html
I have an improved version of the benchmark code which I have been trying
to put the "finishing touches on" for about the past year (!) that sounds
very similar to what you are proposing. Perhaps I should just throw that
into the pot right away? I've implemented about half-a-dozen cases
(textures, lighting, high poly count, transparency, etc.). Each test sits
within the benchmark harness which takes care of logging output and
measuring FPS. Each test must implement a simple interface.
I'm at work right now, I'll dig up the code when I get home and post it to
the list.
Sincerely,
Daniel Selman
At 09:54 PM 9/21/00 -0600, Jim Schatzman wrote:
>I think the idea of having the Java3D community publish a benchmark (or series
>of benchmarks) is a very good idea. Together, we would then compile tables of
>benchmark data and we would be able to get an accurate picture of performance
>on a wide variety of platforms. The activity just might convince one of the
>laptop vendors to put some serious graphics in their box. Certainly it would
>help people decide what to buy or recommend.
>
>I have a benchmark I developed personally which I would be willing to
>donate to the cause. It tests 3D geometry, point array and lighting speeds -
>it has no texture maps or dynamic scene changes.
>
>If this were coupled with benchmarks for texture maps and another for dynamic
>scene building, I believe that we would have covered the bulk of the major
>CPU-cycle
>killers in typical 3D graphics applications.
>
>Is anyone interested in organizing a suite of benchmarks, maybe putting a
>shell around a set of 2-3 with a simple, consistent output? Anyone who
>could distribute such a benchmark suite through their website?
>
>Jim Schatzman
>
>
>
>
>At 12:02 PM 9/21/2000 -0400, you wrote:
> >At Sense8, we developed a 3D graphics benchmark known as Indy3D which was
> >(and still is) fairly widely used for comparing performance across OS,
> >platforms, adapters, configurations, etc. It is OpenGL only, however.
> >
> >The source code (C++ with calls to one of Sense8's products, WorldToolKit)
> >is available for download so someone *could* theoretically create an Indy3D
> >equivalent in Java3D. The model files are in WRL and a Sense8-proprietray
> >BFF format, but I'm sure one of my old friends can convert them all to WRL
> >or some other format.
> >
> >For a look at what Indy3D allows you to do, go to
> >http://www.sense8.com/indy3d/ There is a table there with results, though
> >I don't believe anyone is really keeping it update to do anymore. We
> >released Indy3D in 1997/98.
> >
> >--Mark
> >
> >At 10:05 AM 9/21/2000 -0500, you wrote:
> >>Bob,
> >>
> >>I agree that it'd be nice to have a "benchmark" Java 3D test. But
> >>anyone that knows benchmarking would know that there are a lot of issues
> >>and no one benchmark will cover all the possible factors (ie some video
> >>cards handle textures better than others and a benchmark that doesn't
> >>use textures would be meaningless for reporting video card performance
> >>of handling textures). It sounds a lot simpler than it is.
> >>
> >>I believe at least one person that is active on this list has a
> >>"benchmark" but I think it's rather old.
> >>
> >>This is something I'd like to do but it's way down on the "to do" list
> >>and I'm not sure if I'd ever get around to writing it. I'll give it
> >>some more thought.
> >>
> >>- John Wright
> >>Starfire Research
> >>
> >>Bob Gray wrote:
> >> >
> >> > I'd like to see a performance table for Java 3D listing different video
> >> > cards, configurations, operating systems, frame rate acheived.....
> >> >
> >> > This could be done by first defining and writting a performance test
> >>program
> >> > and making that program available on a web site.
> >> >
> >> > Then, anyone wanting to see how their configuration compares to other
> >> > configurations, they would simple run the *same* program off of the web
> >> > site. The results could be posted here and tabulated on a web site.
> >> >
> >> > The key is that everyone use the same program (or couple of
> programs) and
> >> > report the same information. The program(s) should be completely
> >>automated.
> >> > Just start it and let it run for a while. When its done it would
> display a
> >> > report listing all the info (config, op sys, frame rate, graphics card,
> >> > etc...) You would then only need to post the report.
> >> >
> >> > Anyone interested in making available such a testing program (or two)?
> >> >
> >> > Bob Gray
> >> >
> >> > -----Original Message-----
> >> > From: Jimmy Talbot [mailto:[EMAIL PROTECTED]]
> >> > Sent: Thursday, September 21, 2000 9:47 AM
> >> > To: [EMAIL PROTECTED]
> >> > Subject: Re: [JAVA3D] ATI Rage Mobility 128 + Win2K
> >> >
> >> > Hi,
> >> >
> >> > > This sound about right? Am thinking that probably better to stay with
> >> > > Win98 as the support is better for HW accelaration. Any comments on
> >> > > experiences with Win2K+laptop users out there on how well J3D runs
> with
> >> > > it?
> >> >
> >> > We bought an IBM THinkpad A20p with Rage Mobility 128, Win2k. Although
> >> > RM128 is supposed to support HW T&L, IBM's implementation doesn't.
> That's
> >> > right, here I was thinking I was going awesome performance... But noooo,
> >> > IBM is actually allowed to modify ATI's specs. I've been in talk
> with IBM
> >> > for over two months trying to figure out if IBM's actual hardware
> >> > implementation is at fault, or if it's just a driver issue. Doesn't look
> >> > it's going to get resolved before Christmas.
> >> >
> >> > Jimmy
> >> >
> >> >
> ===========================================================================
> >> > To unsubscribe, send email to [EMAIL PROTECTED] and include in
> the body
> >> > of the message "signoff JAVA3D-INTEREST". For general help, send
> email to
> >> > [EMAIL PROTECTED] and include in the body of the message "help".
> >> >
> >> >
> ===========================================================================
> >> > To unsubscribe, send email to [EMAIL PROTECTED] and include in
> the body
> >> > of the message "signoff JAVA3D-INTEREST". For general help, send
> email to
> >> > [EMAIL PROTECTED] and include in the body of the message "help".
> >>
> >>===========================================================================
> >>To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
> >>of the message "signoff JAVA3D-INTEREST". For general help, send email to
> >>[EMAIL PROTECTED] and include in the body of the message "help".
> >
> >
> >Mark Ferneau 240-462-6262 (cell)
> >Director of Adv. Technology 801-437-4608 (efax)
> >Xtivia Technologies, Inc. 732-469-5954 x629 (NJ office)
> >[EMAIL PROTECTED] 301-279-5703 (home office)
> >http://www.xtivia.com/ [EMAIL PROTECTED] (wireless email)
> >
> >===========================================================================
> >To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
> >of the message "signoff JAVA3D-INTEREST". For general help, send email to
> >[EMAIL PROTECTED] and include in the body of the message "help".
>
>===========================================================================
>To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
>of the message "signoff JAVA3D-INTEREST". For general help, send email to
>[EMAIL PROTECTED] and include in the body of the message "help".
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff JAVA3D-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
