>>This suggests there are several such boards, but the only ones I'm aware
>>of are the Diamond 3D 3000, the STB Velocity 3D, and the ELSA Winner
>>2000AVI/3D -- and the ELSA, I've heard, hasn't even been released yet
>>(due out in January, I'm told). So what other options are there in
>>ViRGE/VX boards? Also, I've noticed that the STB costs ~$60-$70 less
>>than the Diamond in a 4MB configuration, but the specs are identical, as
>>near as I can tell. Why the price difference?
>I don't know. And since my interest in the ViRGE/VX has been helping others
>set up their systems rather than buying one myself, I haven't gone into the
>price/performance tradeoffs.
>If you need hard info about supported boards, talk to the S3V development
>team for XFree86 - they can be reached through email address I posted,
>their homepage at http://www.xfree86.org/, or (sometimes) through this
>newsgroup.
problems with the STB board, though they also didn't explicitly say they'd
heard it works.
I do have another couple of questions for the group on this, though: In
order to get SOME idea of whether this all will work, I pulled the Number
Nine 332 (S3 ViRGE, but not ViRGE/VX) board from my office computer and
temporarily installed it in my home system. I ran xbench on it, and
found that it got 163,363 xstones at 1024x768x(8 bpp) or 92,908 xstones
at 1024x768x(16 bpp), which is rather pathetic for a boad like this --
the ET4000/W32p card I've got now gets 150,707 xstones at 1024x768x(8
bpp) with XFree86 (124,389 with Metro-X). (This is with version 3.2 of
XFree86 for both ET4000/W32p and ViRGE.) Now, I understand that the
current ViRGE drivers are incomplete (xbench results are particularly bad
in the arc and text categories, but 1.5-3x better than the ET4000/W32p
results in other categories), so I've two questions:
1) How much better is the ViRGE/VX than the non-VX ViRGE?
2) How much are the XFree86 ViRGE drivers likely to improve? The
documentation says further acceleration is on the "to-do" list,
but roughly where can we expect to see overall performance
settle -- at 1.1x current levels, 2x, or what?
--
+--------------------------------------------------------------------+
| Rod Smith Author of: |
| http://ezinfo.ucs.indiana.edu/~rodsmith "OS/2 Soundcard Summary" |
+--------------------------------------------------------------------+