Among other things, I think OGD1 could use a couple of killer apps. Obviously, as you say, VGA is a good one, but that's more of a killer app for companies that might want to resell OGD1 boards with some added specializations. But there are some other killer apps we could use OGD1 for until such time as we decide to develop a special product. The whole idea of doing some kind of mega high-end audio thing appeals to me. Of course, we've batted around ideas like this before. But we need to think of things that make us say "I'd pay $1500 to be able to do that specific thing better." They can be relatively niche applications, though. I spoke with someone at Pixar, and unfortunately, hardware acceleration of rendering isn't interesting to THEM because it's so much easier to just add more CPUs to their rendering farm. But there may be other sorts of compute off-load people would find interesting, things that wouldn't be so complex to develop, like cryptography, image/signal processing, and even artificial neural networks. I was talking to one prof at OSU about this. You can fit in some number of networks with back-propagation, but once you have trained your networks, you can reload the FPGA with more than twice as many networks with forward-prop only. Academics may be willing to pay for our hardware to do this.
On Fri, Jun 6, 2008 at 5:30 AM, Dieter <[EMAIL PROTECTED]> wrote: >> I've been following the open-graphics project since early on, and I am >> still excited about it -- in principle, at least. My impression is that the >> project is mostly lacking in reaching out to anyone outside a small core >> group. > > I think a lot of people know about the project, even before the recent > pre-order announcements. It has been mentioned on slashdot and other places > a few times. The question is how to recruit people to become active > participants and contributors, rather than just "let me know when you have > a product to buy"? > >> The website is pretty useful for many (mostly those who are well informed >> already), but it does a poor job at explaining why anyone should care, why >> recent developments (e.g. Documentation released by AMD) have not made it >> obsolete, > > I've been surprised at the lack of discussion on how the AMD/ATI docs > affect OGP. While anyone releasing docs for any chip is a very good thing, > this is *very* serious competition for OGP. While OGP has a few features > AMD/ATI doesn't that are essential to some applications, that probably > doesn't translate into massive sales. On the other hand, AMD/ATI has > video decoding and serious 3D, and those do translate into massive sales. > > Yes, there may be applications other than the FLOSS enthusiast, such > as kiosks. I have no idea how large that market is, or if it will > support the cost of getting a chip fabbed. > >> what the current roadmap is etc. > > Yes, it is not clear to people that the OGD1 is an intermediate step, > and the goal is an inexpensive card that will be available in PCIe. > >> the discussion on the OGP mailing list demonstrated >> a substantial confusion even among those active on the list (e.g. is OGD1 a >> cheap FPGA developent board, and if not, how do we have to qualify the >> statement to make it true?). > > That sounds like me. Despite participating in the OGP list for some time, > I have somehow managed to avoid getting bitten by the FPGAs-are-sooo-coool- > gotta-get-one-to-play-with bug. So I haven't learned all about the various > FPGAs on the market and their pricing. The comments by fpgaprogrammer sounded > at least somewhat knowledgeable, (unlike many of the comments) and seemed > worthy of discussion and perhaps a rebuttal. > >> I, for instance, would be willing to donate money because I believe it's an >> important project. Many others may (_may_, I have no way of knowing) feel >> the same but may have a lower limit (I doubt the blender3d project could >> have collected 100'000 Euro if the lower limit for supporters had been at >> hundreds of dollars). >> >> But donating to the OGP is tricky. Rather than having a prominently placed >> button, the OGP home page contains a note way down the page (you can donate >> to the Open Hardware Foundation), and it's not even a link. The OHF >> homepage, too, has neither a donation button nor a link. It takes 3 clicks >> (FAQs->Donations->How can I donate to the OHF?) to find out that all I can >> do is pledge via a web form. The text telling me about that has apparently >> been posted in late 2006(!) and says "At this time, we are not yet >> donations because we are still finalizing banking arrangements." (how long >> does it take to finalize that, and where's the verb in that sentence?) >> >> I don't think you need more press releases. You need to get your act >> together first -- especially the web site(s). Make it easy for people with >> a casual interest to find reasons to support the project, and make it easy >> for them to support the project -- even if a small donation is all they can >> give. > > Yes, it should be easy for someone to donate money to OHF, and via the > method *they* prefer, within reason. Check, credit card, paypal, ... > What methods are popular in Europe, Asia, Australia, ...? > > In addition to Roger's suggestions, I'd like to suggest that some "firmware" > (for lack of a better word) be available before generating more publicity. > Get the card to *do* something. A minimal VGA would be one obvious choice. > _______________________________________________ > Open-graphics mailing list > [email protected] > http://lists.duskglow.com/mailman/listinfo/open-graphics > List service provided by Duskglow Consulting, LLC (www.duskglow.com) > -- Timothy Normand Miller http://www.cse.ohio-state.edu/~millerti Open Graphics Project _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
