[SOLVED] I need a second opinion(or more) ... Does this look like a fried CPU?

[Spazum]

I popped this CPU out of its socket, because it possible that I inadvertently set the voltage jumpers on this motherboard frying it. The manual states that this board supports P54C/CT/CS/CQS, MMX, Cyrix6x86/6x86MX, AMD-K5/AMD-K6 Socket 7 types.
The voltages range from CORE 2.8v to 3.5v AND I/O 3.5v to 3.3v.

I believe that I set the jumpers for a P54C (3.5v) instead of INTEL MMX (2.9v), because I misidentified the processor.

Zoom closer, and there seems to be a burnt area in the middle that appears melted/faded/corroded. The characters of the ID print in particular.

This should look mostly shiny, but not a blemished/burnt looking spot correct?

What do you all call this prognosis ? I'm hoping i'm wrong. (And please tell me otherwise) LOL ....

Any and all opinions are welcomed and no offense will be taken.

 

[Ralston18]

It appears to me that that CPU got very hot....

 

[Spazum]

It appears to me that that CPU got very hot....

Thanks @Ralston18

 

[TommyTwoTone66]

Did you run it without a cooler? The thing looks cremated. Is this a prank? Did you toss it on a barbecue?

 

[BFG-9000]

The 280nm P55C (MMX) is very sturdy. I've had four of them installed in Socket 5 systems for over 20 years where they've only ever received the single-plane 3.3v intended for 600nm P54C (75-100MHz), a whopping +15% overvoltage. They only have passive coolers on them, too. The only reason the 233MMX chip works at the correct speed in Socket 5 is Intel re-used the old 1.5x multiplier setting as 3.5x for those. Note that Intel only lumped it in with 350nm (like the 120MHz P54CQS and P54CS 133-200MHz) because it's laid out in the same metal pitches (so the density is the same) even if the wiring and transistors are thinner.

That said, I did fry an original AMD K6-233 in the same sort of socket. That 350nm chip was rated 3.2v vCore and 3.3v i/o so I figured straight 3.3v should've been fine. The problem was the VRM circuitry couldn't cleanly deliver the amperage it required so it must have got hit with some nasty power spikes. Oh well, the board still works fine today with a much faster K6-III on an interposer.

I've also seen a 600nm P100 with fried L1 cache before (ran fine with L1 disabled but very slowly as you'd expect) and have no idea how that was managed. Perhaps someone installed it into a 486 adapter and then plugged that into a 5v-only 486 board?

So 3.3-3.5v should not by itself hurt a P55C. Yours is one of the early PPGA packages so the bottom cover is just plastic. Earlier chips than that were all ceramic CPGA which would not show that kind of heat damage