Very cool 😀

However for games I recommend using 3dbench 1.0 and not speedsys.

The other option you have is going for "slower" chips like the standard Pentium 200 and all of the K6-2 versions.

Yeah, I may try a Pentium 200 in a future, if I find one, nevertheless 9.17 is a pretty good slow down, many games will work on that setup....

I don't think the slowed system performs anywhere near a 386DX-40. You can not stick to a single benchmark score. Especially as Mau1wurf1977 already noted Speedsys is not suitable. It mixes CPU and FPU scores to one score and a 386DX-40 with FPU scores about 6.9. Still the performance ratio between CPU and FPU is very different on a P55C compared to a 386/387.
And you also have the local bus to all extension cards that can give in some games twice the performance since graphics accesses are much faster.

I strongly recommend to use some real world scenario like Doom Demo 1.9s to compare performance in games to some oldschool system.
run doom -timedemo demo3 and get the realticks value that is shown afterwards.

My raw guess is that your slowed machine performs more like a 486DX-33.

The black thingy is called a screw terminal.

I particularly like your attention to detail to set the LED display to HI for 233 MHz and 100 for 100 MHz. And as usual, the matching white case to white monitor is most attractive.

I'm also very interested to see what you get in doom, 3dbench, and pcpbench vs. that of a real 386.

On my MMX with all caches disabled and running at 166 MHz I get:

3dbench 1.0: 15.3
Quake timdemo demo1 (all default but no soundcard): 2.1
PCPBENCH/VGAMODE: 3.7

I didn't run Doom when I did these figures, might do this one day though 😀

Wing Commander is certainly a little bit too fast.

Having a fast AGP card vs an ISA ET4000 for example also makes a difference.

RAM also matters. That board has SDRAM but also the older type. SDRAM is a lot faster. On my machine I have one SDRAM that's only 16MB and shows up as EDO in the POST screen. It reduces performance by around 20-30%.

I've got some results:

a) Doom -timedemo: timed 2134 gametics in 22211 realtics

b) PCPBENCH/VGAMODE 640x480: 1.2

c) 3dbench 1.0: 8.8

No idea how to interpret them...

Pingaloka wrote:

I've got some results: […]
Show full quote

I've got some results:

a) Doom -timedemo: timed 2134 gametics in 22211 realtics

b) PCPBENCH/VGAMODE 640x480: 1.2

c) 3dbench 1.0: 8.8

No idea how to interpret them...

Try trixter's TOPBENCH. 😁

New 16-bit-era benchmark project (TOPBENCH)
http://dosbenchmark.wordpress.com/

Cool program Stiletto! First time I see it!
Got some results, don't know how accurate they are, but if so, I'm pretty amazed by the slowdown achieved.
But I'm not quite sure if results are really accurate Stiletto...tried benchmark with Computer L1 + L2 cache on, same setup, and it showed a Pentium II

50mhz 2.0x L1 and L2 Disabled setup = Pentium 100MMX

Any thoughts on the results guys?

Pingaloka wrote:

Any thoughts on the results guys?

8.8 for 3dbench matches an average 386DX I would say. Wing Commander should certainly run really well on your machine 😀

If you're looking for slow, a IDT Winchip is overall slower clock for clock than a Pentium MMX right out of the box without playing with the fsb or multiplier.

sliderider wrote:

If you're looking for slow, a IDT Winchip is overall slower clock for clock than a Pentium MMX right out of the box without playing with the fsb or multiplier.

Thanx good to know....for now I'm going to stick with this setup. Going as down as a 386 33mhz is good enough for me. Now I'm actually going to try play a game, 😀

sliderider wrote:

If you're looking for slow, a IDT Winchip is overall slower clock for clock than a Pentium MMX right out of the box without playing with the fsb or multiplier.

Thanx good to know....for now I'm going to stick with this setup. Going as down as a 386 33mhz is good enough for me. Now I'm actually going to try play a game hehe this is just the perfect setup for my needs 😀

That's a beautiful case.

Thanx hwh, yes it really is a nice case!