VOGONS


Reply 220 of 495, by KEC

User metadata
Rank Newbie
Rank
Newbie

Hey guys I love the forum and the benchmark suite.. I was going threw my computers and doing a full cleaning since I use to smoke around them and they're pretty nasty. Stumbled across this thread and started benching my systems, which has really got me tweaking and tuning everything.. I've found a nice bit of untapped power in each one..

I've added 3 machines so, working on the 4th now and I'm noticing some trends..

1. VIA's DOS performance is horrible

2. Nvidia's DOS performance is horrible, esp with a VIA chipset.. Appears Geforce 6 series and above are some how limited in DOS.

3. Voodoo cards rock for DOS and scale incredibly well.. My V3 3000 AGP scores Identical to an ATI X800 and kills my Nvidia cards.
It's seems to suggest the V3 is keeping up with a 1g PIII which is cool considering it came out with CPU's in the 400-500mhz range
and only supports AGP 2x. Also the V5 5500 is identical to the V3 3000 in DOS. It acts like only use the first of the two GPU's
is being used so maybe it needs windows drivers to unlock the 2nd.

4. I'm getting odd results in DOOM. On every machine I'm gunning for the top spot, and I always win Quake but lose DOOM.
The V3 seems to have hit a ceiling at around 510-520 ticks.. Not sure if it's some odd limit of the V3 or if it's VIA biting me in
the ars:) I'll know more as I get into stronger comps...

5. In DOS the Via Apollo Pro 133A vs the BX is land slide for the BX. My P-III 1g with PC133 cas 2 is struggling to keep up with BX
boards at 850 mhz and PC100. I may have a BX board in the closet, but not sure I have a PIII that's fast enough to make it shine
or if it supports the 1g at a lower clock.. I'll look into it...

Anyways, great forum and thx for bringing life back into my old comps.. Been a long time since I stayed up late searching for
drivers and tweaks:) As much as I love new computers, sometimes they just aren't as much fun.. Granted fast and reliable is a
wonderful thing, but sometimes a guy needs to get down and dirty in search of the Demon Tweak:)

Reply 221 of 495, by gandhig

User metadata
Rank Member
Rank
Member

Yeah i'm at the receiving end with it. Let me know if you find a 'Demon Tweak' apart from the memory interleave and pci latency one for the via chipset. Keep the benchies coming.

Dosbox SVN r4019 + savestates Build (Alpha)
1st thread & the only one related to the forum(?)...warning about modern-retro combo
Dead, but, Personal Favourite
Replacement for Candy Crush...Train the Brain

Reply 222 of 495, by kithylin

User metadata
Rank l33t
Rank
l33t
KEC wrote:

2. Nvidia's DOS performance is horrible, esp with a VIA chipset.. Appears Geforce 6 series and above are some how limited in DOS.

Actually, what you're seeing is the Voodoo3 series have an actual 'performance' 2D/3D engine for DOS games designed in to it. Where as by the 6xxx series nvidia cards, they were shifting far away from dos (no one used it any more by then) and the focus was purely on DirectX performance. Therefore these newer cards did not have a powerful 2D / DOS-3D engine even built in to the card any more, and will not run fast for DOS games on any system you put them in. I have a geforce 6800 ultra and it's rather slow in almost every system I've put it in for DOS games. Where as.. for example, a geforce 2 ultra us nearly +8x (or more) faster in Phil's benchmarks for dos applications.

It's just to do with how cards are designed, newer cards are crap for older setups in general. Period-appropriate hardware is always going to be many magnitudes faster for older game performance, I've found.

Reply 223 of 495, by elianda

User metadata
Rank l33t
Rank
l33t
KEC wrote:

1. VIA's DOS performance is horrible

I can't see this from the bench. All the later VIA chipsets you refer to I guess are in the top score range. Even the KM400 with integrated Unichrome graphics is close to the P4 2.4 GHz. It is much more about the used graphics card.
And about NVidia, you actually see that there is no considerable gain anymore for cards newer than a GF4.

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 224 of 495, by armankordi

User metadata
Rank Member
Rank
Member

Gonna unload big time on this bitch. Anything you haven't covered in 386-486 land, i'll take that bitch.

IBM PS/2 8573-121 386-20 DOS6.2/W3.1
IBM PS/2 8570-E61 386-16 W95
IBM PS/2 8580-071 386-16 (486DX-33 reply) OS/2 warp
486DX/2 - 66/32mb ram/256k cache/504mb hdd/cdrom/awe32/DOS6.2/WFW3.11
K6/2 - 350/128mb ram/512k cache/4.3gb hdd/cdr/sblive/w98

Reply 225 of 495, by RacoonRider

User metadata
Rank Oldbie
Rank
Oldbie
armankordi wrote:

bitch

Perhaps you meant "bench"?

Phil, could you please organize stock and overclocked lists? The calculation for new entries has stopped.

P.S. My Am5x86 just made the best among Am5x86 in Quake 😁 12.7 FPS and still some say SIS is a slower chip set.

Reply 226 of 495, by vetz

User metadata
Rank l33t
Rank
l33t
RacoonRider wrote:

P.S. My Am5x86 just made the best among Am5x86 in Quake 😁 12.7 FPS and still some say SIS is a slower chip set.

Where were you in the other thread? I also know that the SiS chipset is generally faster in SVGA resolutions.

3D Accelerated Games List (Proprietary APIs - No 3DFX/Direct3D)
3D Acceleration Comparison Episodes

Reply 227 of 495, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++
RacoonRider wrote:

Phil, could you please organize stock and overclocked lists? The calculation for new entries has stopped.

Should work again 😀

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 228 of 495, by RacoonRider

User metadata
Rank Oldbie
Rank
Oldbie
Mau1wurf1977 wrote:
RacoonRider wrote:

Phil, could you please organize stock and overclocked lists? The calculation for new entries has stopped.

Should work again 😀

Great! Thank you!

vetz wrote:

Where were you in the other thread? I also know that the SiS chipset is generally faster in SVGA resolutions.

What other thread?

Reply 229 of 495, by easy_john

User metadata
Rank Member
Rank
Member

Thank you for this thread and benchmark results.
Very helpful, when I change 486 vlb mobo and need to find, that it works not slower, than should be.
Added 2 lines to bottom of the table with my results.

Pentium2 450/256mb/4gb/ati rage 128+voodoo2/SB awe32 8mb+db50xg/GUS PnP 8mb/TB Tropez 2mb
486 DX2-66/32mb/8gb/tseng4000 2mb/SB 16+WB/GUS 1mb/LAPC-I
286 12mhz/4mb/512mb/Vga 1mb/SB 2.0+Covox
PegasosII G4 / Amiga 4000 / Amiga1200 / Amiga 600

Reply 230 of 495, by sunaiac

User metadata
Rank Oldbie
Rank
Oldbie

My P133 is not in the charts.
This makes very cute cats very sad 😢

R9 3900X/X470 Taichi/32GB 3600CL15/5700XT AE/Marantz PM7005
i7 980X/R9 290X/X-Fi titanium | FX-57/X1950XTX/Audigy 2ZS
Athlon 1000T Slot A/GeForce 3/AWE64G | K5 PR 200/ET6000/AWE32
Ppro 200 1M/Voodoo 3 2000/AWE 32 | iDX4 100/S3 864 VLB/SB16

Reply 231 of 495, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++
sunaiac wrote:

My P133 is not in the charts.
This makes very cute cats very sad 😢

The charts are just too much time to create, so I'm sorry 😵

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 232 of 495, by easy_john

User metadata
Rank Member
Rank
Member

No need a charts.
Better sort table by socket, then mhz, then fps - to easy compare your system with another.

Pentium2 450/256mb/4gb/ati rage 128+voodoo2/SB awe32 8mb+db50xg/GUS PnP 8mb/TB Tropez 2mb
486 DX2-66/32mb/8gb/tseng4000 2mb/SB 16+WB/GUS 1mb/LAPC-I
286 12mhz/4mb/512mb/Vga 1mb/SB 2.0+Covox
PegasosII G4 / Amiga 4000 / Amiga1200 / Amiga 600

Reply 233 of 495, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Yes you can do this yourself by clicking on VIEW > LIST. You can then sort and filter. E.g. only Intel or only VIA chipsets.

The other option is exporting the DATA to Excel and working from there 😀

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 234 of 495, by sunaiac

User metadata
Rank Oldbie
Rank
Oldbie
Mau1wurf1977 wrote:
sunaiac wrote:

My P133 is not in the charts.
This makes very cute cats very sad 😢

The charts are just too much time to create, so I'm sorry 😵

Told my cat you were sorry.
Seemed to make him rejoice.

(maybe you should remove them alltogether no ?)

The main problem I have reading the main table is taht there's too much information before the scores.
Can't the four first columns be the scores ?
(I know it's not easy and many ppl already proposed an organization for all this data, so just another input 😀 I appreciate the huge work behind it !)

R9 3900X/X470 Taichi/32GB 3600CL15/5700XT AE/Marantz PM7005
i7 980X/R9 290X/X-Fi titanium | FX-57/X1950XTX/Audigy 2ZS
Athlon 1000T Slot A/GeForce 3/AWE64G | K5 PR 200/ET6000/AWE32
Ppro 200 1M/Voodoo 3 2000/AWE 32 | iDX4 100/S3 864 VLB/SB16

Reply 235 of 495, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Is that on a Full HD monitor or on a 800 x 600 retro PC 🤣

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 236 of 495, by kithylin

User metadata
Rank l33t
Rank
l33t
Mau1wurf1977 wrote:

Is that on a Full HD monitor or on a 800 x 600 retro PC 🤣

I have no problems viewing the tables in google docs format on the website in a 1400x600 16:10 monitor

Reply 237 of 495, by RacoonRider

User metadata
Rank Oldbie
Rank
Oldbie
kithylin wrote:
Mau1wurf1977 wrote:

Is that on a Full HD monitor or on a 800 x 600 retro PC 🤣

I have no problems viewing the tables in google docs format on the website in a 1400x600 16:10 monitor

I have those problems with 1280x1024 19" LCD monitor which I still consider the optimal choice for anything. Still, there's Excel and you can actually memorize/put down which string number is for which computer. Ta-da!

Reply 238 of 495, by RacoonRider

User metadata
Rank Oldbie
Rank
Oldbie

Made a long video card benchmark run with my new build-in-progress, AMD 5x86 -133 on Acer AP43 (SIS 496/497).

Matrox Athena showed the expected poor DOS performance; however I expected better from V2200. S3 videocards show impressive scores yet I was confused when 64V+ turned out to be slower than 64, given that 64 has 60ns EDO ram and 64V+ - 45ns. S3Trio 3D/2X has the best score of all S3s, yet I expected it to slightly outperform Millenium II and ET6000. Well, it didn't, despite SDRAM and being manufactured in 2000.

The top scores are achieved with Matrox Millenium II, S3 Trio 3D/2X, ET6000, Matrox Mystique, but ET6000 did not work in DOS text mode, so Doom score is unknown. We see almost equal scores: at this point, the CPU is no longer bottlenecked by the videocard.

P.S. I'm adding Millenium II score into the list. Are the other scores worth adding? It would create confusion if I added V2200 and Athena scores there; S3 scores are more or less identical; ET6000 is inconsistent.

Attachments

  • video.png
    Filename
    video.png
    File size
    16.58 KiB
    Views
    1663 views
    File license
    Fair use/fair dealing exception
Last edited by RacoonRider on 2014-05-03, 08:55. Edited 1 time in total.

Reply 239 of 495, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Very nice!

Totally up to you what / if you want to add anything 😀

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel