On 5/24/05, Tiago Mikhael Pastorello Freire <[EMAIL PROTECTED]> wrote:
> Let me join the i-am-an-enduser-too crowd.
> 
> You know, I am able to shed up to US$ 200 for a FPGA just to foster the
> effort, but looks like it is going to cost more than that. I can still donate
> some change for the effort regardless of buying hardware.
> 
> I want to share my thoughts with the developers, maybe you could estimate the
> feasibility of this alternative plan.
> 
> I see a world trend in having portable devices like mp3 players/portable game
> devices/PDA/smartphone cobbled together.
> 
> As such, I am just wating to see a really good linux solution come by, which
> has not happened so far. I am lurking int his list, as well as on the Gumstix
> project (I have already posted about them before), and I really think that a
> synergy between both projects would be most welcome.
> 
> I would like to hear from the developers if they have already figured out how
> much simpler will the embedded part be, if the embedded FPGA version will be
> significantly smaller, and thus cheaper.

There is no embedded FPGA, and the ASIC and FPGA versions will have
the same functionality.

> 
> Maybe this could be the way of having Open PDA kits for linux geeks out there?
> Depending on the answers, maybe it could be easier to do, even though it is
> smaller than the PC market. Care to spend a few brain cycles on this
> possibility?

The OGP design would be at the high-end for hand-held systems, but it
would likely make a very good PDA GPU.  If someone wants to design an
open PDA, we'll cooperate with them fully in integrating our GPU with
their design.

> Another thing which comes to my mind is:
> We have computer chips using a socket, so that we can replace the CPU without
> changing the motherboard. Would it be possible to do the same with the chip,
> having two parts, a base daughter board, and either a slot in which we can
> put either the FPGA or the ASIC (future upgrade!!), or if it is not possible
> to fit the chip in a slot-like format, make a granddaughter board with the
> chip.

Custom sockets are not cheap.  Plus, most of the ASICs we deal with
are ball-grid-arrays--going to pin-grid-array for a socket would
result in an increase in packaging cost.  Also, note that subsequent
iterations of the design may have changes in pin-out.

> IMO it could be more attractive upgrade-wise, even if each part (board and
> chip) are individually more expensive. This way I would buy my base card with
> FPGA, then buy the ASIC upgrade when it comes by, and afterwards upgrade my
> base board to the video capture one.
> 
> And everyone would be happy :)

The objective of the first iteration of the OGP is to minimize
development time and cost.  We need to get SOMETHING out the door
before we start getting fancy.

_______________________________________________
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