> I really like this Steam idea.  If we could get them to participate in the
> project, that would be phenomenal.  Should we wait until I've had time to
> code the reference stream processor?

I think it is useful to have something tangable to show potential
partners.  I think the OGD1 should be sufficient. In addition, we
need a good roadmap plus some ideas of how to make a partnership work
well for both parties.
-----------------------------------
Roadmap rev 0.000001:

simulator + foo + bar + massive amounts of testing -> chip design + chip docs
chip design + big money -> mask
mask + money -> chips

In parallel:

work -> board design + board docs
board design + money ->  boards

boards + board docs + work -> FLOSS device drivers (BSD, Linux, Plan 9, ...)
boards + board docs + FLOSS device drivers -> Framebuffer cards to sell
boards + board docs + FLOSS device drivers + chips + chip docs + work
 -> FLOSS GPU software
boards + board docs + FLOSS device drivers + chips + chip docs +
 FLOSS GPU software -> GPU cards to sell

Also need:
website
corporate entity to deal with money
sales channel(s) DIY? Partner(Steam/whoever)? Amazon/Newegg/whoever?

Initial marketing is easy: announce availability to slashdot, phoronix,
FLOSS mailing lists.

Obviously this needs a <BLEEP> of a lot of work.  More of a gant chart
than a roadmap.
-----------------------------------
Partnership ideas:

Perhaps there could be a vanilla version chip and a "Steam Powered"
version with extra features of some sort. (I have no clue what
these features might be. But Steam might have ideas on features that
would be useful in videogames. Hmmm can we partner with the protein
folding people? Features to cure cancer faster would be far more
useful than better videogames.) To save mask costs make the dies all
the same. When packaging, an enable-steam pin either gets connected
or not. This might be as simple as a seperate power pin for that
circuitry. The vanilla chip uses less power, the Steam version has
more features, more shaders or whatever.

And of course the possibility of designing the chip so that defective
parts can be deactivated and sold as a lower spec chip, like they do
with multicore CPUs.

Steampunk marketing theme. (Or does Steam already do this? I have no
clue, not into videogames.)

Keep in mind that there are lots of people that are not into
videogames. And I assume that not all videogamers are into Steam.
So avoid things like all cards get bundled with Steam stuff
that non-videogamers don't want to have to pay for.

> They have gotten help from the graphics card
> vendors, but they've also had complaints.

Do we know what these complaints are?

> Steam might be sensible to not listen to us until we
> have a complete set of masks for a complete GPU in 32nm process technology.

If we can get to the point where we have the masks, we can surely
go the rest of the way without them.

> Nothing is stopping us from getting there besides a bunch of distractions.

and $2M or so for the masks.

If we have people with strong business deal negotiating skills who
aren't skilled at designing GPUs or boards, they could go off and
think about partnering stuff awhile and report back.

> First, we do the things we can do now, with the resources we
> have, and then we use those to achieve something else.

Anyone have a good feel for how many framebuffer cards we could
sell for applications that don't need a GPU? POS, ATM, digital signage,
stockmarket, servers, grandma, CNC machines, medical, air traffic control,
9-1-1, ... There are tons of applications once you get away from
protein folding, 3D CAD, animators, videogames and such. But not as easy
to sell to as the FLOSS crowd, or videogamers. Imagine all the hoops to
jump through to sell to air traffic control.  But we could sell a lot
for server consoles.
_______________________________________________
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