Sounds like the motherboard doesn't know the CPU and can't run without knowing it properly. My Aopen AP43 behaved that way before I updated the BIOS with my 5x86 (it was fine when configured as a DX4, but using the 4x multiplier made the computer to hang)

I had a roughly similar problem. I have this soyo board wich is similar but not the same as those i could find on stasons. When i tried my d2-80 it would always triple the clock (75-100-120), no matter what i jumpered. After reading some amd datasheets i found that there where some amds which use another pin for the clock-command "CLKMUL" (which is often writeback/through for other CPUs). I had to do the wire trick 😀 as the pic shows to get a gnd signal for that pin.

Also you should run something to check the "real" frequency of the cpu like CHKCPU. When i had 25mhz fsb the afore mentioned dx2-80 showed as dx-80 in POST but obviously ran at ~75mhz.

Greetings

It happens alot on a few of my boards when running any cpu over 100mhz it will show it as 100mhz as the bios. doesn't know any higher. But using chkcpu will show the cpu running at the real speed, don't always believe the post screen

It's funny, on the old 5V 486 motherboards I can get just about any of the 486 going using a VRM. The BIOSes on the newer boards really seem to make life hell.

Treeman was right. It shows 100 MHz on the POST screen, but all the benchmarks I ran from Phil's Dosbench CD actually show the CPU running at 120 MHz. And after re-seating the IDE controller it also sees the hard drive now.

Still it is giving me headaches, because with the 120 MHz CPU it is actually SLOWER than with the 80 MHz one, according to most of Phil's benchmarks. The only two showing it faster is the Speedsys one, which just measures the CPU speed and the DOOM Benchmark with high details shows it at least on par around 20 FPS. All the others show it slower than the same system with the 80 MHz CPU. I also ran it with the 80 MHz and a different graphics card (ISA ATI Wonder instead of a Diamond Stealth 24VL VLB) and it is faster than that variant.

So i AM seeing the improvement I would expect from going from the ISA VGA card to the VLB one. But going from the 80 MHz CPU to the 120 MHz doesn't do anything noticeable. Unless all those Benchmarks are almost exclusively depending on the graphics card and the CPU doesn't matter, but i always thought especially DOOM was taxing on the CPU. Any advice would be appreciated.

I have a AMD DX2-80 and mine is actually a rebranded DX4, maybe yours runs at 120MHz too ? Is it identified as a DX2-80 by benchmarks ?

when I was using my amd 586 133 at normal speed benchmarks would show it correct, when I clocked it to 160 suddenly my score fell to like 30 from 100+ speedsys still shows it as 160 tho and quake score shows the improvement. Somebody said in a post somewhere once that some of the benchmarks don't identify the AMD chips correct and get confused. Run a quake or doom time demo I usually only use quake from 80mhz to 120mhz im taking a guess should be about 1-2fps improvement, then you can safely say you are running at 120mhz

I just saw you got doom there, there seems to be a slight improvement but I have not too much experience in doom. scores

Deksor wrote:

I have a AMD DX2-80 and mine is actually a rebranded DX4, maybe yours runs at 120MHz too ? Is it identified as a DX2-80 by benchmarks ?

How does one know it is rebranded? As is stated above mine would run at 120, but i never did any benches.

Well I learned that by mistake. I configured it wrongly once, and saw "486DX4" and indeed it was running at 100MHz, so that's a DX4 in disguise x)