As everyone here is probably aware, Intel has good open-source support
for their integrated graphics, including their new GMA X3000.  The
performance doesn't match the top end ATI and NVidia chips, but it's
a reasonable midrange solution.

Of course, as an integrated solution, it doesn't help those of us that
want to use a motherboard a different chipset, or that want multiple
cards.  (The Intel chips support two monitors each, provided that you
don't want them to act as a PCIe x16 bridge, but some people actually
want more than two monitors.)

After looking at the data sheet on the G965 chip (which contains the
GMA X3000), it occurred to me that it might be possible to put the
G965 on a card with an FPGA (probably Spartan 3) to act as a PCIe-to-FSB
bridge.  The main technical issue was whether the GMA X3000 actually needs
the FSB cache coherency protocol, and I've been told that it does not.

The cost of goods of such a graphics card would be low, and the
software effort to support it is minimal since X.org already supports
the chip in the normal (integrated) configuration.  The only obvious
changes that might need to be made to the X.org code would be address
translation (because the on-card memory would be mapped from address
0 from the card's POV, but from a very high address from the CPU's POV),
and possibly code changes to support multiple cards.

I considered pursuing this on my own, but decided that I don't quite
have the resources to do it myself.  I could probably handle the
RTL design of a PCIe-to-FSB bridge, but the PCB would have to be
designed by someone experienced in very high speed layout and with
appropriate signal integrity simulation tools.  Debugging may require
a rather expensive logic analyzer with PCIe and Intel FSB preprocessors,
which would cost quite a bit even to rent.

Possibly a first stab at it could use plain old PCI rather than PCIe,
which would be useful for a multi-card solution.

I suspect that if cards were developed and minimally tested, that the
X.org and Intel software people would help with any necessary software
changes, though I certainly haven't gotten any commitment from them
(nor asked for one).

Anyhow, it's not my intention to distract anyone from the OGP
development, but if anyone thinks this might be a worthwhile thing
to build, I'd be interested in discussing it.

Eric Smith

_______________________________________________
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