Return to BSD News archive
Received: by minnie.vk1xwt.ampr.org with NNTP
id AA5705 ; Fri, 01 Jan 93 01:53:15 EST
Xref: sserve comp.unix.bsd:9454 comp.os.linux:20829 comp.unix.pc-clone.32bit:848
Newsgroups: comp.unix.bsd,comp.os.linux,comp.unix.pc-clone.32bit
Path: sserve!manuel.anu.edu.au!munnari.oz.au!spool.mu.edu!umn.edu!math.fu-berlin.de!informatik.tu-muenchen.de!roell
From: roell@informatik.tu-muenchen.de (Thomas Roell)
Subject: Re: ET4000/W32 and VESA VL-Bus
In-Reply-To: hasty@netcom.com's message of 28 Dec 92 22:37:27 GMT
References: <1hloneINNfnn@cbl.umd.edu> <1992Dec28.103211.12677@diomedes.robots.ox.ac.uk>
<1992Dec28.190316.22747@Informatik.TU-Muenchen.DE>
<1992Dec28.223727.3569@netcom.com>
Sender: news@Informatik.TU-Muenchen.DE (USENET Newssystem)
Organization: Inst. fuer Informatik, Technische Univ. Muenchen, Germany
Date: Tue, 29 Dec 1992 19:18:33 GMT
Message-ID: <1992Dec29.191833.1401@Informatik.TU-Muenchen.DE>
Lines: 79
>Both Jon and I ran xbench using the Actix GraphicEngine32 (801) and we both
>got 64k xstones. However, I ran xbench at 1024x768 45MHz interlace.
Then S3 has with the 801 a pretty good DRAM design. I would now post
some 928 results, if I would be allowed to, but I can tell you that
the difference to the 801 is QUITE BIG.
>What we are trying to compare here is the effect of running a DRAM based
>card at lets say 71MHz vs. 45 MHz interlace which you have stated in your
>previous posting that running at a high dot clock rate the performace of a
>DRAM based will be significant worse.
I said, that this effect depends on the resolution, the total memory
bandwidth, the internal screen refresh queue. If you are below a
border of requirements for the screen refresh, you wouldn't notic this
effect at all (like 640x480 v.s. 800x600). If S3 came around this,
great. I would like to see the WHOLE xbench output to analyse this
more in depth, and some second sources that also ran BOTH , 75MHz and
45MHz dot-clock at the SAME machine. This would give intresting
results. If S3 really managed to get that high performance out of
DRAM, then congratulations. But still my comments VRAM vs. DRAM apply.
>Can you be kind to post your Utra Pro benchmark results?
TOTAL 348878 lineStones
TOTAL 58865 fillStones
TOTAL 49077 blitStones
TOTAL 5037233 arcStones
TOTAL 185281 textStones
TOTAL 111307 complexStones
TOTAL 100895 xStones
>Also, is 64k xstones is that what your are getting?
Nope, that's the number flying around for the XS3 server. X386 1.3 is
quite a bit faster. Don't have benchmarks for a 801 card around, but
here is one for the Orchid Farentheit1280, the slow one with the 911,
the first S3 based board really shipped ...
TOTAL 267658 lineStones
TOTAL 23236 fillStones
TOTAL 18637 blitStones
TOTAL 2760416 arcStones
TOTAL 138531 textStones
TOTAL 46470 complexStones
TOTAL 44306 xStones
>According to ACTIX's Director of Marketing the S3 928 is going to list for
>about $350. The ACTIX GraphicEngine32 is listing for $199.
>It will be interesting to find out from different manufacturers what are
>they planning to sell the 928 for :-)
I would say that depends on how much more over the list price they
have to pay S3 to get 928 chips at all ...
Will be intresting. Have another benchmark number for people who like
really hot boards ... This is the Nth Engine/250, a VRAM based board
with a 82C481 on it... Maybe the 928 will be faster ;-)
TOTAL 320838 lineStones
TOTAL 74961 fillStones
TOTAL 54213 blitStones
TOTAL 4722855 arcStones
TOTAL 262281 textStones
TOTAL 115620 complexStones
TOTAL 118481 xStones
- Thomas
PS: Sorry for the commercial numbers; wanted just to make clear how
dimensions are between VRAM and DRAM based system.
--
-------------------------------------------------------------------------------
Das Reh springt hoch, e-mail: roell@sgcs.com
das Reh springt weit, #include <sys/pizza.h>
was soll es tun, es hat ja Zeit ...