Fake KU80486SX-33 SX684 (tested and verified)

[debs3759]

I've tested the bulk of my 486 and about 2/3 of my Socket 7 chips. The only oddity I found (odd because I wasn't expecting it...) is that one of my SX684 is a fake.





The top one has a reset ID of 0422, stepping B0. The bottom one has an ID of 0428, stepping cB0. As SX683 has an ID of 0428, and the only other source I have looked at with stepping info shows it should be 0428, the 0422 must be fake. I'm not really surew what the point was for making a fake QFP chip on a PGA adapter - surely whoever did it could have profited more by faking a PGA chip!

This is why I like to keep different examples of a chip that look different, until such time as I can determine whether they are or not

[debs3759]

SX683 is a 486DX25 in the same package. In this case, it is evident that one or the other is a fake, as the same sSpec will always have the same CPUID value. As Intel used consecutive Spec numbers in many cases to i9ndicate the same core running at a different speed, and as the SX683 I ahve tested has ID 0428, it makes sense to initially assume that is correct.

I then refered to a list I have mirrored locally (I can't remember the original link, unfortuantely, as I would like to contact the author, David L. Beem, about it...), and the 0428 was confirmed by his list.

Also, the one I believe (and state) to be fake has print that is quite heavily raised from the surface of the chip, whereas the other has the regular, flat print.