Baoran wrote:amadeus777999 wrote:Overall a 486 is pretty slow, even on a good board, when compared to its successors. […]
Show full quote
Overall a 486 is pretty slow, even on a good board, when compared to its successors.
I'll see if I can test a 100mhz Pentium(fsb 50mhz) against the already established scores tomorrow.
I had a i486(2x50 mhz - all settings max) running against a PII at 2x50mhz on an P2L97 and it only faired well in Doom...
iDX2-100 -- performance relative to a 100mhz PII
Doom(demo3) 95%
UDoom(demo4) 83%
Blood 70%
3DBench 51%
Quake 34%
Albeit above said is not directly related to your performance demands, a 3D accelerator still needs to be fed.
100Mhz pentium 2?
I was more like wondering if a 100Mhz 486 might match 60Mhz pentium when pentium 60Mhz seems to be the minimum system requirements for most of those games that he mentioned that would not be playable on a 486.
It "matches" it unless the software is floating point oriented.
The Pentium66 that I measured against the DX2-100 was slightly slower in Doom - speed was around 95%(no matter if it was a "light" or "heavy" demo) whereas in Blood it reached 107% - both had a Matrox Mystique installed.
I have no 33mhz based 486 systems, but my guess is that a normal DX4@33mhz fsb is quite bit slower than the "pumped" DX2-100, which should further accentuate the Pentium's efficiency.
I also had a Pentium60 paired with a Voodoo2 but it was nothing more than a "dang, it works" project. GLQuake felt very laggy which was to be expected.
Update:
I tested a Tekram i430TX based board(+EDO ram and same MatroxMystique) @50fsb(+P55C@125mhz) and scaled the scores down to 100mhz. Compared to this setup the iDX2 100 is able to perform at...
85% in Ultimate Doom(demo4)
87% in DoomShareware(demo3)
58% in Blood
This board is a bit slower especially when compared to a SiS5591 based one, so the difference may be more pronounced there.