Good scores. Shows that the board and maybe BIOS settings do play a role. The main thing I was looking for is show the differences (rather than absolute scores).
You get them as well, so when it comes to DOS Gaming on ISA, the choice of ISA VGA card makes quite a difference. Would love to know what the ET4000 scores in your system!
Would getting a crappy trident or oak pci card be useful in slowing down a K-6 II for older DOS games, (along with the disabling cache trick)?
Compared to AGP? Yes, somewhat. An ISA card would make a larger impact however.
In my case the fps went from ~ 150 (AGP) to ~ 40 (ISA).
So the ISA bus moreso than a crappy card. ISA slots are slim though. Im thinking I should have gotten the Asus P5A-B or FIC 503+ AT board, instead of the ATX board. They have 3 ISA slots to the normal Super7 ATX boards 2 ISA slots.
Depends on the board, my Iwill SS7 board is ATX and comes with 3 ISA slots.
But a PCI card should also slow things down, not sure what's a crappy one though...
Underclock the PCI bus maybe?
Im wondering how these 3DBench scores are relevant to 2D cards.
wrote:Im wondering how these 3DBench scores are relevant to 2D cards.
The results speak for themselves!
Im wondering how these 3DBench scores are relevant to 2D cards.
Um... because 3DBench measures 2D speed. Or at least certain specific aspects of 2D speed.
I think you're not understanding what the program does. It's not like modern 3D benchmarking... back when 3DBench came out, 3D was done with software rendering and displayed through 2D cards... 3D accelerators didn't come along until later.
Ha! Ha!
So early 90s 3D games wont be accelated by the fancy cards that came in the late 90s....as they werent written to 3D standards or directly for the 3D accelerator chips.
wrote:Would love to know what the ET4000 scores in your system!
Well, apparently not so good... exactly the same as Elsa Winner 1000 - here we have:
- Diamond SpeedStar VGA (ET4000AX, 512KB): 50.0
- Diamond SpeedStar 24X (WDC, 1MB): 58.8
Would an extra 512KB improve things for ET4000AX? I intend to look for those anyway...
Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
Maybe back in those DOS days it just sounded "cooler". Remember, noone saw the whole 3DFX revolution coming 😉
I mean, they used to call the newer machines "Multi Media PC's". Wow, what a name!
wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
not really... it IS doing 3D graphics, using the methods available at the time
Plenty of old DOS games used to render polygons for their graphics, and did it all on the CPU... of course they were often flat coloured surfaces and not textured, but it was still 3d graphics 😀
wrote:wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
not really... it IS doing 3D graphics, using the methods available at the time
Plenty of old DOS games used to render polygons for their graphics, and did it all on the CPU... of course they were often flat coloured surfaces and not textured, but it was still 3d graphics 😀
Like Stunts!
Oh man, the countless hours it waisted!
And indeed, it was 3D and no textures whatsoever. Grass was just 1 large green area 😁
But the gameplay...absolutely fantastic!
wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
You need to play 4D Sports Driving to see the 3D we are talking about. 😉
wrote:Like Stunts! Oh man, the countless hours it waisted! And indeed, it was 3D and no textures whatsoever. Grass was just 1 large gr […]
wrote:wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
not really... it IS doing 3D graphics, using the methods available at the time
Plenty of old DOS games used to render polygons for their graphics, and did it all on the CPU... of course they were often flat coloured surfaces and not textured, but it was still 3d graphics 😀
Like Stunts!
Oh man, the countless hours it waisted!
And indeed, it was 3D and no textures whatsoever. Grass was just 1 large green area 😁
But the gameplay...absolutely fantastic!
I believe it's the same game we are talking! We pratically answered at the same time! When i think of old 3D these names just pop right out, and the other one is 4D Sports Boxing.
Yes stunts was an awesome game!
Especially with the track editor, made so many awesome tracks...
Another one I used to play was called 'Strike II', you had a rear third person view of a ship and you flew around destroying things, trying to earn money to buy upgrades etc
If you are squeamish, don't prod the beach rubble.
wrote:wrote:Does anyone else find it strange that a program called 3DBench is used to test VGA cards that can't do 3D?
not really... it IS doing 3D graphics, using the methods available at the time
Plenty of old DOS games used to render polygons for their graphics, and did it all on the CPU... of course they were often flat coloured surfaces and not textured, but it was still 3d graphics 😀
If it's all done on the CPU then isn't 3DBench more a benchmark of the CPU than the video card? Would it matter at all what video card you use if all the work is done by the CPU?
The graphics card does have an effect- just check the 3dbench database thread on these forums 😀
Often merely swapping graphics cards can make the same systems results change dramatically.
It does seem to be a more CPU-oriented benchmark but even if all the calculations are done on the CPU, often the graphics card can't push the pixels to the screen fast enough to keep up and so brings the score down.
The difference is even more noticable at higher resolutions as video performance, rather than raw CPU power, becomes the bottleneck. Using PCPBench at 640x400 or above, the results differ even more wildy just by swapping video cards, even though all the graphics are still rendered by the CPU.
