On 11/26/06, Dag Bakke <[EMAIL PROTECTED]> wrote:
Are the quoted "guesstimated" performance numbers for OGD1/OGA for the
graphics card running with the FPGA at its core, or with the FPGA code
converted into an ASIC?

With the FPGA.  The ASIC would be much faster.

 From what I have been able to understand of the discussion on the m/l,
it should be possible to run the FPGA at 150-200 MHz, which matches the
quoted numbers.

That's the idea.  We've done other things with the 3S4000, and
although it's quite a challenge to hit 200Mhz, 100 is generally quite
easy, and 150 isn't overly difficult.  I'm shooting for 150.

Thus:
Is it possible to provide an educated guess about the performance of a
card running with the FPGA converted into an ASIC?
(Yes, I realise the uncertainty associated with estimating the
performance of a next-generating product based on a not completed
current-generation product...)

I think Howard had guessed that, in general, it would typically be
about twice as fast.  You can get better if you pay more.  Not
everything will scale the same since the ASIC will be a different sort
of cell technology.  We'll end up having to tweak things again, so who
knows.  But I wouldn't be surprised if we hit a core clock of 300Mhz
and a memory clock of 500Mhz or better.

And by the way: someone mentioned the 7600gs having dual-link DVI. The
7600gs does have dual-link DVI, but I am not sure if both ports on
(all?) dual-port cards are dual-link enabled.

I think having two full dual-link ports will be a good selling point.
_______________________________________________
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