VOGONS


Reply 22 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
Darkman wrote:

just one question , in Quake 2 , which texture quality did you use ?

Default settings. So whatever that is 😊 The only setting I changed was render mode to default OpenGL and the resolution.

YouTube, Facebook, Website

Reply 23 of 97, by Darkman

User metadata
Rank Oldbie
Rank
Oldbie
philscomputerlab wrote:
Darkman wrote:

just one question , in Quake 2 , which texture quality did you use ?

Default settings. So whatever that is 😊 The only setting I changed was render mode to default OpenGL and the resolution.

cool , sadly I dont own forsaken that cant be benchmarked, but I can benchmark a few other games like MDK2 (which is one of the first games to really favour the Geforce due to having hardware T&L)

Reply 24 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++

Really, Unreal and UT are the main ones, seeing they use Glide.

The version I used is the DVD Unreal Anthology.

YouTube, Facebook, Website

Reply 25 of 97, by Darkman

User metadata
Rank Oldbie
Rank
Oldbie
philscomputerlab wrote:

Really, Unreal and UT are the main ones, seeing they use Glide.

The version I used is the DVD Unreal Anthology.

not sure if that makes a difference, I use the original boxed version (not even the GOTY version) but with the latest official patch , so it should be the same really.

Reply 27 of 97, by meljor

User metadata
Rank Oldbie
Rank
Oldbie

interesting, would like to see that acer board vs the Tusl2-c

asus tx97-e, 233mmx, voodoo1, s3 virge ,sb16
asus p5a, k6-3+ @ 550mhz, voodoo2 12mb sli, gf2 gts, awe32
asus p3b-f, p3-700, voodoo3 3500TV agp, awe64
asus tusl2-c, p3-S 1,4ghz, voodoo5 5500, live!
asus a7n8x DL, barton cpu, 6800ultra, Voodoo3 pci, audigy1

Reply 28 of 97, by F2bnp

User metadata
Rank l33t
Rank
l33t
philscomputerlab wrote:

Yea, doubt it makes much of a difference, if any.

AFAIK, Unreal Gold used an updated engine, straight off of UT. So, if you were to compare Unreal Gold vs Unreal vanilla, especially on lower resolutions, I believe vanilla would have a slight edge. But it is of no matter, since the framerates would already be skyhigh. 😀 I'll be benchmarking in a while and post the results later today. I expect my VIA board to be trailing the Intel boards at 640x480 and 800x600, but I think on higher resolutions there should parity between the boards and results within margin of error.

Reply 29 of 97, by Darkman

User metadata
Rank Oldbie
Rank
Oldbie

benchmarking in Windows 2000 complete , Win2K SP4 was used , along with the standard 3DFX drivers.

GLQuake

640X480 - 337.1
800X600 - 234.6
1024X768 - 154.3
1280X1024 - 102.6
1600X1200 - 72.7

Quake 2

640X480 - 182.2 3DFXGL - 242.3
800X600 - 145.1 3DFXGL - 202.6
1024X768 - 105.4 3DFXGL - 145.6
1280X960 - 73.0 3DFXGL - 100.6
1600X1200 - 42.1 3DFXGL - 66.5

Unreal

640X480 - 153.97
800X600 - 137.14
1024X768 - 108.75
1280X1024 - 73.07
1600X1200 - 51.80

UT99

640X480 - 99.44
800X600 - 98.55
1024X768 - 94.40
1280X1024 - 74.40
1600X1200 - 54.08

For Quake 2 I ran both Default OpenGL and 3DFX OpenGL , since the Default values seemed very low for some reason.

interesting to see the diffrences between Win98SE and Win2000 , the CPU seems to be taxed more at lower resolution in W2K , whereas the results are almost identical at higher resolutions (could be because Win2K puts a bigger load on the CPU in general , or 3DFX's Win2K drivers probably being pretty early)

Some other benchmarks I ran

Quake III (all maxed settings at 1024X768) - 77.4 fps

3Dmark01 - 2219

Final Reality
2D Image Processing - 13.85
3D Performance - 8.28
Bus Transfer Rate - 7.01
Overall Score - 9.76

Now for the performance of the U160 Seagate Cheetah drive in this machine

1f4d707b-bb0e-4cad-b63f-bbc3cc9ac15d_zpshj1yhbrp.png

so far this is it.

Reply 30 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++

Very nice results Darkman. That SCSI drive is very impressive I must say.

meljor wrote:

interesting, would like to see that acer board vs the Tusl2-c

So would I, could you send me one please? 😊 🤣

F2bnp wrote:
philscomputerlab wrote:

Yea, doubt it makes much of a difference, if any.

AFAIK, Unreal Gold used an updated engine, straight off of UT. So, if you were to compare Unreal Gold vs Unreal vanilla, especially on lower resolutions, I believe vanilla would have a slight edge. But it is of no matter, since the framerates would already be skyhigh. 😀 I'll be benchmarking in a while and post the results later today. I expect my VIA board to be trailing the Intel boards at 640x480 and 800x600, but I think on higher resolutions there should parity between the boards and results within margin of error.

Cheap way to get Unreal Gold is the GOG version. If it's on sale. Use a reverse installer to unpack it, remove the filed for the glide wrapper and you're good to go 😀 I also got the DVD version. For some reasons there are a ton of UK sellers, selling old PC games, NIB for little money. Got quite a few that way 😀

YouTube, Facebook, Website

Reply 32 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
F2bnp wrote:

Phil, is the Forsaken benchmark running at 16bit color or 32bit? Ran it once at 640x480x16, but Vsync seems to be enabled and I only got 75FPS, any idea on how to get rid of that?

I think 16 bit colour.

Yes, the driver has V-Sync on for DX games. You need to run the overclock tool (it's an official tool from 3dfx). It adds another page to the driver and here you can disable V-Sync for DirectX games.

YouTube, Facebook, Website

Reply 33 of 97, by Darkman

User metadata
Rank Oldbie
Rank
Oldbie

hmm, I went ahead and quickly tested a couple of games in Win98SE , and the numbers were actually identical to Win2K

meaning there is something slowing down this PC by a bit, will have to find out what is it.

Reply 34 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++

I wouldn't drop any sleep over this. Apart from using the tightest memory timings, what else can you do?

What chipset driver did you use? I went with an older one. 2001 driver from here: http://www.philscomputerlab.com/intel.html

YouTube, Facebook, Website

Reply 37 of 97, by Skyscraper

User metadata
Rank l33t
Rank
l33t

This is a very nice build 😀. I wish I soon find a cheap Voodoo5, they always end up going for more than Im willing to pay.

The PIII 1400-S sure is popular among us VOGONers 😁

New PC: i9 12900K @5GHz all cores @1.2v. MSI PRO Z690-A. 32GB DDR4 3600 CL14. 3070Ti.
Old PC: Dual Xeon X5690@4.6GHz, EVGA SR-2, 48GB DDR3R@2000MHz, Intel X25-M. GTX 980ti.
Older PC: K6-3+ 400@600MHz, PC-Chips M577, 256MB SDRAM, AWE64, Voodoo Banshee.

Reply 38 of 97, by PhilsComputerLab

User metadata
Rank l33t++
Rank
l33t++
Skyscraper wrote:

This is a very nice build 😀. I wish I soon find a cheap Voodoo5, they always end up going for more than Im willing to pay.

The PIII 1400-S sure is popular among us VOGONers 😁

Yes they cost a small fortune. I got mine from an Australian collector. He sold up everything on eBay, so I grabbed a V4 and V5 and he did free shipping 😀 It wasn't cheap, but the prices aren't going to come down and you can wait forever.

F2bnp wrote:

Timings/BIOS Settings etc.

Could someone upload or direct me to the latest version of ATTO benchmark that will work with Win9x? 2.41 and 2.47 won't run 🙁.

Hmm 2.41 should work. Here is an older screenshot I took:

Alternatively try Disk Speed. While it doesn't test write performance, it does test random reading and also access time.

FL8gJpg.png

Attachments

  • Filename
    DSsetup.exe
    File size
    689.99 KiB
    Downloads
    51 downloads
    File license
    Fair use/fair dealing exception
  • Filename
    atto_disk_benchmark_241.zip
    File size
    232.97 KiB
    Downloads
    73 downloads
    File license
    Fair use/fair dealing exception

YouTube, Facebook, Website

Reply 39 of 97, by Darkman

User metadata
Rank Oldbie
Rank
Oldbie

well , Ive tried several chipset drivers for both Win98SE and WIn2K.

Win98 did benefit from the older drivers, lower resolution results are higher. On the other hand, Win2K shows no improvement , and was actually slightly worse in performance, meaning I settled on the 2002 drivers as a compromise.

As far as BIOS and memory timings , the RAM in this is already at CL2 , and while this Intel board is rock solid in terms of stability, it has relatively few tweaking options.

though of course, if the games already run at 200+ fps, a couple of frames here or there arent going to make a difference.