Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!elroy.jpl.nasa.gov!decwrl!sgi!shinobu!odin!westworld.esd.sgi.com!erik
From: erik@westworld.esd.sgi.com (Erik Fortune)
Newsgroups: comp.sys.ibm.pc.hardware
Subject: Re: Mentor & Trident VGA cards - which one is the best ?
Message-ID: <1991Apr10.230915.18440@odin.corp.sgi.com>
Date: 10 Apr 91 23:09:15 GMT
References: <1991Apr3.071713.6290@fel.tno.nl> <3725@d75.UUCP> <1991Apr10.100931.27295@Informatik.TU-Muenchen.DE>
Sender: news@odin.corp.sgi.com (Net News)
Reply-To: erik@westworld.esd.sgi.com (Erik Fortune)
Organization: Silicon Graphics Inc.
Lines: 53

In article <1991Apr10.100931.27295@Informatik.TU-Muenchen.DE>,
roell@informatik.tu-muenchen.de (Thomas Roell) writes:
>I know the ET400 *very* well but I don't know of any VGA board that uses this
>feature (if it really exists). But on the other hand the overhead of band-
>switch in less than 1% (if you have two segments).
Hmm.  I'll try to dig up my reference for it.  I think it
was in the "programmers guide to ega, vga, and svga" published
by Brady.   I'll look for a more complete reference.

It's too bad that nobody uses the feature (if it exists) --
it was the ET4000 feature that most leapt out at me as
I scanned the book.

>There is also a 64k banking mode, which is much easier to handle. The 128k
>segments would break applications that use a VGA and a monochrome adapter.
>In this mode you have only to modify one register for segment switching
instead
>of two for the 128k segments.
Agreed.
The only time I'd recommend using the 128k segment register mode is for 
a 4-bit mode that needs more than 64k.  In other words, for any mode that
doesn't require you to touch the segment registers.  

The part that I thought was particularly bizarre was the way you switch
between 64k and 128k mode.  I don't remember exactly which register it
is, but writing to one register sets it to 128k mode, and reading from
the register toggles the mode (or the other way around).

>Yes but the problem is ONLY with scrolling. Perhaps with some chaching tricks
>the overhead could be reduced. For example with the ET4000 you have one
>read operation followed by one write operation. This means the processor has
>to wait for getting videomemory access. The ET4000 reads one line (32 bits)
>at once and buffers it. But cause of the following write this buffer is once
>again flushed. But the write will be buffered in a write-buffer internally.
>If the TVGA8900 has the same architecture the sequenece of reads and writes
>will make full use of such internal buffers. And if you read only small blocks
>of data that fit into one line of the 386 cache there might be an additional
>improvement. Concluding I assume that the overhead of the TVGA8900 of the
>ET4000 in scrolling would be about 40 to 60%. All other operations shounld
>have the same speed.

I agree that we could play some games to minimize the overhead of the
register swapping, but it annoys me that we'd have to.   It means a moderately
large amount of special-purpose code which could have been easily avoided.

>This statement is INEXACT. If the ET4000 & TVGA8900 WOULD support a 8 plane
>mode they would be equally fast. But they both (?) use a 8 bit packed pixel
>mode for displaying 256 colors.
Oops.  When I said "8 plane" I meant "8 bit" (in X terms 8-bit Z format).
I don't deal with too many XY format devices (in fact, the vga is the only
XY device I've used).

-- Erik