GeorgeMan wrote:

33 results uploaded!
I saw no point in trying P133-K6/300 on the 2 newer motherboards, if anyone else wants just say it. 😀
It is also pointless for me to bench more PIII/Celeron (366-933MHz) as they are pretty much too fast for these DOS benchies.

There's never anything "too fast" for this stuff. I put my Core2Duo + HD4890 in there near the top end of the results at one point.

GeorgeMan wrote:

33 results uploaded!
I saw no point in trying P133-K6/300 on the 2 newer motherboards, if anyone else wants just say it. 😀
It is also pointless for me to bench more PIII/Celeron (366-933MHz) as they are pretty much too fast for these DOS benchies.

Also added some s370 results. Now I only have a 486 100-120MHz VLB left and a 386SX.

There's a whole lot of results, mainly thanks to GeorgeMan! Phil, could you please sort them?

RacoonRider wrote:

There's a whole lot of results, mainly thanks to GeorgeMan! Phil, could you please sort them?

Done! Thanks for reminding me 😊

Please stickify this! Very useful 😀

Entered my info from my Dell Dimension 4100. Looks like I'm on the low end of fps for Doom and Quake. I blame the Nvidia GPU BIOS on this card. (GeForce FX 5500)

I also find it weird that I'm the only person with a Pentium III running at 866MHz. Stock setting for it and I can't change it if I wanted to. BIOS won't let me touch much.

I can bench some of my PS/2s if it would help (the PS/2E or model 33 has a 486SLC @ 50 Mhz which is quite a unique processor, and the PS/2 95 has a pentium 60 with insane cache running on synchrostream). So both of those processors will perform unusually faster than their 'regular' counterparts.

Normally they would use XGA, but I have a VGA 'MCA' card coming for the 95, and I can always throw an ISA VGA card in the model 33. Unless XGA would be preferred to be used for the testing for a curiosity factor...

Let me know,

Basic info on the PS/2E: http://www.walshcomptech.com/ps2/mod33.htm
Quick overview of Synchrostream: http://www.walshcomptech.com/ohlandl/complexes/complex.html

Silver/blue chip to the right of the purple ceramic intel cache controller is what performs synchrostream operations:

Cache chips are the five chips above both the heatsinked P60 and cache controller (and there are five more underneath for a total of 256 KB: for a strange amount of approximately 25 KB per chip).

raymangold wrote:

I can bench some of my PS/2s if it would help (the PS/2E or model 33 has a 486SLC @ 50 Mhz which is quite a unique processor, an […]
Show full quote

I can bench some of my PS/2s if it would help (the PS/2E or model 33 has a 486SLC @ 50 Mhz which is quite a unique processor, and the PS/2 95 has a pentium 60 with insane cache running on synchrostream). So both of those processors will perform unusually faster than their 'regular' counterparts.

Normally they would use XGA, but I have a VGA 'MCA' card coming for the 95, and I can always throw an ISA VGA card in the model 33. Unless XGA would be preferred to be used for the testing for a curiosity factor...

Let me know,

Basic info on the PS/2E: http://www.walshcomptech.com/ps2/mod33.htm
Quick overview of Synchrostream: http://www.walshcomptech.com/ohlandl/complexes/complex.html

Silver/blue chip to the right of the purple ceramic intel cache controller is what performs synchrostream operations:

Cache chips are the five chips above both the heatsinked P60 and cache controller (and there are five more underneath for a total of 256 KB: for a strange amount of approximately 25 KB per chip).

Strange indeed! As Phil stated earlier, new entries are always welcome!

RacoonRider wrote:

Strange indeed! As Phil stated earlier, new entries are always welcome!

Okay what I'll do is test the model 95 with an MCA VGA card, and then I'll leave the model 33 with its XGA (which is VGA compatible), just for the sake of bizarreness as there aren't any XGA entries yet. One thing I did forget to mention, the model 33 is completely fanless and uses a passively cooled PSU! A very nice touch if you want a mini pizza box pseudo-486 that doesn't generate any noise.

I'm going to bench a couple of systems myself - oddly, when using your menu system, the benching doesn't work with Doom on my Vaio PCG-FX601, but it works fine when I run the batch file directly from the Doom subdirectory.

EDIT: Vaio and my HP Vectra VLi8 added.

Added 28 more results (reaching a total of 70 results), mainly from my slot1 sets, including many 440BX results! 😁 😁

I have one more i440BX mobo to test and then I'm finished.

Wow what a marathon! That's awesome.

mau1wurf1977, could you let me know what settings you are using for the Quake timedemo?

When I use your Quake 1.06 benchmark, I get 12.4 fps. When I use my Quake 1.06, I get 14.4 fps (timedemo demo1).

feipoa wrote:

mau1wurf1977, could you let me know what settings you are using for the Quake timedemo?

When I use your Quake 1.06 benchmark, I get 12.4 fps. When I use my Quake 1.06, I get 14.4 fps (timedemo demo1).

Default settings. Didn't change a thing.

This phenomena could lead some people to think their systems are faster or slower than they really are. I wonder if it is just my system which displays this effect. Can anyone else with a 486 compare their stand-alone Quake SW 1.06 results with the Quake results in Phil's VGA package? A 15% discrepency is very curious.

Errr. How do I run timedemo on quake?

I beleive Quake requires you to have DOS 5.0 or above installed, or Windows 9x. I do not think you can run Quake from a boot disk even if you have the required files already on your HDD.

1) Install quake from the installer file by running install.bat.
2) run quake from DOS by typing: Quake
3) Once in quake, start a new game
4) Once the game starts, type: `
5) This will bring up the quake command console.
6) Now type: timedemo demo1
7) And immediately type: `
😎 This will close the command console and run the quake timedemo without the console taking up half the screen.
9) When the timedemo is finished, the command console will show you the frame rate.

But you need to use Quake files not part of Phil's VGA package. Even if I run Phil's Quake files independent of the batch bench file, it still indicates the decreased frame rate. I am not sure why. I thought all the 1.06 shareware versions were the same.

EDIT: The checksums of the 4 files in this zip compilation are the same as the ones I used. http://www.moddb.com/games/quake/downloads/qu … -shareware-106/

Ok there is something weird going on.
Running the original quake files (made from a copy of someone elses disks)
I am seeing much _LOWER_ performance (8.9 FPS)

And with Phil's bench I am seeing 11.0 FPS!

-edit-
I will try the linked version
The version you linked is also 8.9FPS

The opposite effect - very strange indeed. Are you running the benchmarks from within DOS or within Windows' DOS prompt?

Is it possible that you have sound enabled? Try removing the sound card or disabling sound in Quake (there is some command for this, but I don't recall right now).

If I remember correctly Phil's benchmark run at fullscreen without UI, while default settings have the UI enabled. This explains the 15% performance difference.

According to Phil in an earlier post here on Vogons were I pointed this out it was set to fullscreen to be more heavy on faster machines (same applies to Doom).