🚀 go-pugleaf

In article <1admlnINN6aj@ecoole.soph.cs.ubc.ca> t8c192@soph.cs.ubc.ca (Michael Angelo Relova) writes:
> >I'm wondering if in order to take full advanatge of a local bus video card,
>application has to recognize that the data path to the card is
> >32 bits as opposed to 16, 8 bits.  Applications I'm talkinare
> >normal VGA or SVGA type of DOS apps.  I'm aware that regardless of the width
>
>You should be getting a MUCH faster video rate... but to the answer 
> to your question, your programs can't tell whether or not you have 
> a local bus or not... I mean it may be able to detect it with a benchmark 
> prog or something, but as far as other progs a
>re concerne, it's just another VGa card
>

Well, I've tried various shareware benchmark programs (BTW, my system is a
GHost ET4000 Local Bus 486-33).  This has been the result:

3dbench: This one does some 3d animations in VGA mode and the calculates the
	 Ave frames per second.  My result: 14.7.  Other users have claimed
	 about the same (or better) using a 386-40 ISA with an ATI or Diamond
	 VGA Card.

vidspeed: This runs some vga blocks on the screen the tells you: Processor,
	 CPU Speed, VGA Bus Width, and Frames per second.  It correctly
	 ID's bus width at 32, and processor 486.  FPS comes out to ~60, and
	 the docs say ~15 is very good!

vgaspeed: This runs thru all the video modes and writes/reads data from the
	 VGA display.  It consistently gives >9000 Bytes/ms for writing in
	 all display modes.  Seems pretty fast to me.

winspeed: This tests Windows specific performance of processor and video.
	 Note: I'm running this on OS/2 Winos2.  It tells me that both
	 CPU and Video performance are between OK and Good (falls between
	 386-33 and 486-25).

So, I'm confused.  2 of the tests seem to indicate very good performance and
2 not so great.  *is* there a definitive test for vga speed?

-edg-