http://en.wikipedia.org/wiki/Am5x86
Commonly overclocked to 160MHz (4x40) - while some may reach 200, this requires a motherboard and RAM that is stable at 50MHz FSB

NB. some are 3.3V, others 3.45V

http://www.ocbase.com/ OCCT, or maybe Prime95

Ive read good things about this board today. How can i find out is my RAM stable? I use 16MB FPM RAM currently (by Texas Instruments Z417400)

Stojke wrote:

What is the best way to test AMD 5x86 P75 overclock stability?
I have obtained an Biostar MB-8433UUD-A Version 2 that can push the clock to 40 MHz and 50MHz. And i have 4 AMD P75s (2x ADW and 2x ADZ version).

Sadly the battery is as dead as a battery can be, and its soldered, so i will need to remove it. But unlike one person who had problems because of that mine passes the boot phase.

Prime95 25.11?

Prime95 and OCCT are going to be extremely slow on a 486 class CPU. It's also mainly a memory and FPU test and 486s are more about integer performance. I'd probably try looping Doom or Descent somehow.

I'm waiting to hear about actual 200mhz success. I never got one to even POST.

I am about to try that with the 4 chips i have 😀
First 160 than 200. But i cant find my copper core cooler 😒

The copper core is not going to help, slap on a tec on the cpu and cooler.

Well, 3 of them run flawlessly at 160MHz. None managed to start at 200MHz.
One of them wont work at 160MHz.

swaaye wrote:

Prime95 and OCCT are going to be extremely slow on a 486 class CPU. It's also mainly a memory and FPU test and 486s are more about integer performance. I'd probably try looping Doom or Descent somehow.

I'm waiting to hear about actual 200mhz success. I never got one to even POST.

Those programs are made to test modern CPUs with tightly integrated FPU and the FPU almost always destabilizes before the ALU
On 486s the FPU and ALU are only really integrated as far as being on the same die and having an interconnection
and thus the ALU can be unstable with a stable FPU or a unstable FPU and a stable ALU. I think you would need to test them independently

Prime95 and similar really are not suited well for pre-pentium checking

I would recommend running the Quake 1 DOS benchmark for 3 rounds. If that is successful, try for an install of Win98SE.

There was one user here who got that Biostar board running at 200 MHz and it seemed stable enough to install Win98SE and run several benchmarks without fail. He had to use 5 V to the CPU though. You will have the most success if you limit the number of memory sticks to 1 and use double-banked 256 KB of L2 cache. Start off by setting your cache timings to 3-2-2-2 and your FSB-to-PCI bus to 2/3. The read wait state for the RAM may also need to be set to 1 WS.

I have never had success running an X5 at 180 MHz, let alone 200 MHz. I was able to get one to POST at 180 MHz, but it was not stable.

I would forget about 200MHz. If you have to run the CPU at 5V you're probably going to fry your L1 cache.

Quake is very good at finding instabilities.

I used it to test stability with overclocking an AMD K6-III+.

😊 just saw your post feipoa...

Using SuperScape (3DBench2) all processors score 74.6FPS when running at 160MHz with S3 DX 2MB.
Pro ghetto bench blaster: http://puu.sh/arJqq/b6c4dae521.jpg

Awesome test bench! We need more test benches 🤣

It can get a bit ugh when my hair gets into the fan 🤣

Are p75's that are stable at 160MHz rare?

Stojke wrote:

Are p75's that are stable at 160MHz rare?

I don't know. Mine did it without isses. Same motherboard as yours. I believe I entered the results in the OC section of the VGA processor database.

Stojke wrote:

It can get a bit ugh when my hair gets into the fan 🤣

Are p75's that are stable at 160MHz rare?

No, they are quite common. 180 and 200mhz is another game 😀

Stojke wrote:

Using SuperScape (3DBench2) all processors score 74.6FPS when running at 160MHz with S3 DX 2MB.

This is low, what is their score at 133 ?

Every ADZ-marked X5 chip I've tested has run fine at 160 MHz. If you are having issues with an ADZ chip, it may be related to your CMOS timings.

The timings are defaulted as you wrote (due to dead battery).
I have two ADZ and both work at 160MHz, while one ADW doesn't.

The score at 133MHz is 70.5FPS.