Does the spec say "Q0520"? That is not found in CPU-World database ...

I would think the earliest possible proto pentiums would be 1992.

"1990 date code" on a (C)1993 chip?

H3nrik V! wrote on 2024-06-04, 12:32:

Does the spec say "Q0520"? That is not found in CPU-World database ...

The spec code should be above the goldcap. Unfortunately it is not readable.
Below the goldcap there's normally the date code. I'd read it as "C0520", hence "C"=manufacturing site, "0"=1990 and "52" = week 52. Of course I could be wrong on this but that's how I'd interprete it according to my knowledge. Or, "0"=2000, but that wouldn't make much sense either.

https://www.cpu-world.com/forum/viewtopic.php?t=17713

Definitely not a 1990 date code, with a 1993 copyright. Also, 2000 is too late. I'm not convinced this CPU has a date code on it.

OK, Q0520, convinced 😉
Looking at the picture at cpu-world.com the date code indeed is above the goldcap. Could've sworn it was vice versa. Age is hitting...
So this part is authentic and sure enough a nice & rare collector's item! Lucky bidder.
Thanks all for clearing this up!

What I’ve always found pathetic was Intels pension to always screw early adopters with non-upgradable sockets.

Socket 4 p60 wasn’t even released and Intel was already planning socket 5 and 3.3 volts.

They should have released 1 pentium socket with a voltage selector so folks wouldn’t get sandboxed

rmay635703 wrote on 2024-06-04, 18:56:

Socket 4 p60 wasn’t even released and Intel was already planning socket 5 and 3.3 volts.

Haha, they had to.
They even had to overvolt their 66 MHz Socket 4 Pentium slightly from 5.0V +-5% to 5.15V +-5%
And it got so hot!

That thing shouts fake to me. Or perhaps it is some sort of memento, made way later out of dead chips.

I haven't worked at Intel, or designed any chips, but I've seen prototypes. You do not put a logo and other brand markings on a prototype. Laser engraved serial on the package - maybe. But no copyright. What for, this is not going to leave the lab.
Frankly an early prototype would not be packaged like this at all, how do you access the silicon with the lid already soldered? Unless it was already so far down the testing phase that the core was working without issues and this was a socket compatibility test article - in which case it would probably already have proper spec code and other type markings. Simply because such chips would also be sent to mobo makers.

rmay635703 wrote on 2024-06-04, 18:56:

What I’ve always found pathetic was Intels pension to always screw early adopters with non-upgradable sockets.

^penchant ... was meant I presume.

Engineering sample (ES), not a prototype.

https://www.intel.com/content/www/us/en/suppo … processors.html

rmay635703 wrote on 2024-06-04, 18:56:

What I’ve always found pathetic was Intels pension to always screw early adopters with non-upgradable sockets.

Socket 4 p60 wasn’t even released and Intel was already planning socket 5 and 3.3 volts.

They should have released 1 pentium socket with a voltage selector so folks wouldn’t get sandboxed

Technically a socket 4 to socket 5 interposer should be possible

Would need level shifters 5v to 3.3v for signaling if I recall correctly and 3.3v supply for vcc from there you might even be able to seperate out vcc2 and get k6 in there.

The bios is obviously going to have a stroke with that, but this all seems very possible.

But with how rare socket 4 boards are, motivation is pretty low.