Luck of the draw, usually 3.6 is the highest you want to go or the chip can get gimped where it only runs on 4v and eventually wears out.

I’ve seen a few run 5v their entire life’s so it’s worth the risk if you got the chip cheaply enough that you can replace if it gets flakey

rmay635703 wrote on 2022-09-09, 13:26:

Luck of the draw, usually 3.6 is the highest you want to go or the chip can get gimped where it only runs on 4v and eventually wears out.

I’ve seen a few run 5v their entire life’s so it’s worth the risk if you got the chip cheaply enough that you can replace if it gets flakey

So does that mean that at least it wouldn't fry the cpu immediately when turning on the computer?

In small increments, increased voltage will result in accelerated degradation of chip after some time (it's called electromigration), which may happen after years or decades of prolonged usage. More voltage you add - less time is needed for it to happen. But in some cases, 3v 486 chips can work for prolonged periods of time on 5v, if sufficient cooling is provided, without any signs of degradation, i.e. not being stable on specific frequency with specific voltage which was stable before.

From my point of view, 4v is still "safe" to use on all 3v 486. That's only 16% increase over stock voltage (3.45v). More voltage is getting into dicey territory.

By mistake I was using a DX4 100 with 5v for a long time and recently I did undervolting tests with that same processor and it is totally stable at 2.7v so at least in my case I can say that these processors are quite resistant.

My cpu is actually the 133Mhz version but it has been marked as 100Mhz part. It has 16kb of internal cache and it can be jumpered between 3x and 4x. Maybe quality control found issues when it was manufactured and it was labeled as 100Mhz and 8kb cache because of that. All tests show 16kb internal cache though.