I tried running a first generation K5-100mhz at 120mhz-2x60,system boots,but hangs-up after.

Not very well and they do need very good cooling for any overclocking at all, will need a board that can handle the cpu well. Use a copper cooler or find a modern cooler that can be modded to mount. As for the compound don't use the cheap stuff, overall limit for most is around 120 mhz and often less. With aggressive cooling and a bump on the volts you could do 120mhz stable and might post at 125 to 133mhz.

The K5 tends to be near its limit at whatever bin AMD labeled it with. 133 MHz is probably near the limit of the architecture considering how the 133 MHz PR200 was barely released. I do have a PR200 and IIRC mine gets unstable at around +10 MHz.

If you cooled one with liquid nitrogen it might go faster thanks to the laws of semiconductor physics. 😉

The fastest I was able to get a K5-PR166 to run stable at is 120 MHz. This is a whole 3.5 MHz faster than rated. I was able to run the majority of the 686 benchmarks at 125 MHz, but some of the more demanding ones would not complete.

Strange that even at .35 micron they couldn't get it to run faster.

I've overclocked my K5-PR166 to 120mhz. It had no problems with the benchmarks I ran. Have not tested it further, but I'll give it a try. I have two chips of this rating, so I*ll test both 😀

oh god, i wished that they would at least run at 83*1.5=125. 😢
and they are rated at 3.5volts, which is the maximum for most boards already, so how can there be any possibility of overvoltaging?

K5 isn't a good overclocker. They ran it to it's limits which was one of the reasons why they desperately needed a sixth generation part to replace it and bought out Nexgen. Nexgen's 686 part was further along in development than AMD's. It would have been another 2 years before AMD's own 686 hit the market and by then Intel would have moved on.

noshutdown wrote on 2014-06-04, 00:56:

oh god, i wished that they would at least run at 83*1.5=125. 😢
and they are rated at 3.5volts, which is the maximum for most boards already, so how can there be any possibility of overvoltaging?

My old PCCHIPS m571 had undocumented voltages above 3.5 volts

The k5-pr133 I had could overclock to 120mhz at 4ish volts

Hmm mm, reading this all I started to appreciate my pr166 that apparently runs at 133mhz even more. True though that I've never tried it outside DOS. It may or may not be fully stable... And I haven't tried it over 133mhz either.