Download the VGA benchmark from here. It has an easy to use menu and does it all for you. You can then compare the settings with the database:

Phil's Ultimate VGA Benchmark Database Project

Regarding your questions.

3dbench comes in 2 version. Did you use version 1.0c? This is the one you should use for faster computers.

PCPBENCH runs in high resolution unless run with the /VGAMODE option. So maybe one card is much faster in high resolution?

Results for machine 1 seem all very logical to me. They are in accordance with my results in a similar system and Phil's results in the benchmarks database.

For machine 2 Doom & Quake results seem again correct and logical for the configuration.
For 3DBench & PCPBench though (based on doom & quake scores) I'd expect something around 165 and 55 accordingly, so definitely something's wrong with this numbers... Most probably the benchmark results themselves and not your system of course.

yhep, ver 1.0c

ok. i wasn't looking at phils results I was looking at some other page. comparing it to that machine 1 does look in the correct range. I have no idea why pcpbench and 3d bench give such wacky readings on machine 2 though. I didn't mess with any options just ran the programs same way on both machines.

soviet conscript wrote:

well, I downloaded some benchmark tests and finally decided to benchmark a few DOS machines for fun but after running a few prog […]
Show full quote

well, I downloaded some benchmark tests and finally decided to benchmark a few DOS machines for fun but after running a few programs the results seem really weird to me so tell me if these look right

machine 1 = Pentium 133, Virge, 128MB RAM, 512kb L2 cache

3dbench = 87.2
DOOM (v1.9 no sound) = 49.73 FPS
PCPBENCH = 34.1
Quake = 28.6 FPS

machine 2 = Pentium 200MMX, Virge/GX, 128MB RAM, 1MB L2 cache

3dbench = 1522
DOOM (v1.9 no sound) = 81.54 FPS
PCPBENCH = 26.3
Quake = 48.3 FPS

a few things seem off to me, first off the 2nd machine is faster yhea but the 3dbench score difference seems a bit much to me. also doing some comparisons the doom scores seem a bit low for both systems in comparison to comparable systems. why is the PCPbench score so much lower for the faster machine?

Are you sure your are using the same version of 3dbench? looks like one is giving a score and the other is giving framerate
The ViRGE is probably holding you back particularly on the second system

yhea, these machines are more geared to compatability rather then speed so thats why I went the virge route

ok, I must of been really tired last night. I figured it out. ran everything again and now it looks better, ha, for 3dbench I actually was just missing the "." thanks guys.

machine 1 = Pentium 133, Virge, 128MB RAM, 512kb L2 cache

3dbench = 87.2 FPS
DOOM (v1.9 no sound) = 49.73 FPS
PCPBENCH = 34.1
Quake = 28.6 FPS

machine 2 = Pentium 200MMX, Virge/GX, 128MB RAM, 1MB L2 cache

3dbench = 152.2 FPS
DOOM (v1.9 no sound) = 81.54 FPS
PCPBENCH = 60.2
Quake = 48.3 FPS

Looks about right

on a side note...and this seems like a pretty dumb question but how do you transfer over all files and directories in DOS 6.22?

on the previous 2 PC's I have a dos shell program and transferred the benchmark programs from my ZIP drive that way. on a few other PC's I do not so when I try copying the directory over from the ZIP drive it only copys the files and not the directories. and yes I did use the xcopy command but it still will not copy the directories and its a lot to do 1 at a time

ex.

xcopy e:\bench c:\bench

not working, only copying the files in the bench directory but not the subdirectories.

What I do is go to the folder on the source drive

So let's say I want to copy all games. I go to C:\GAMES

Then I go the destination drive. Let's say I want to copy to D:\BACKUP so I go to D:\BACKUP

Than I type C: which gets me back to C:\GAMES and I type

XCOPY *.* D: /s

I guess I'm just old fashioned, but I prefer to copy using the full path names, just to be on the safe side.
With your example, that would be:

xcopy c:\games\*.* d:\backup /e

I prefer /e to /s as it also included empty folders, such as an empty save game folder or such.