ADA3700AEP5AP

[Dome]

Don’t worry about the price; I did not pay much for it precisely because of the questionable markings. That is part of why I find it interesting as a collector.

But your last reply actually reinforces my skepticism about the simple counterfeit explanation.

You say these unmarked C0 ES chips were “very common” at Athlon 64 launch, and that fakers could have laser-marked them later.

But if unmarked authentic ES material was already available, what exactly would be the logic of taking a genuine collectible ES, adding suspicious non-standard markings, using two different fonts in one line, and choosing an anomalous AP OPN that immediately raises questions?

A profit-driven counterfeiter would normally reduce suspicion, not create it.

That is the part I still do not find logically explained.

And regarding silicon lottery: I have now completed a 7-hour Prime95 run at 2.0 GHz using only 1.25 V at ~34.5 °C. Together with 10-hour validation at 2.4 GHz / 1.55 V, that suggests at minimum a very strong sample.

I am not claiming that proves a secret AMD project. But it does make me question reducing everything to “obviously fake” as if no uncertainties remain.

Also, calling the counterfeit explanation “fact” seems too absolute. It still depends on assumptions about what a faker supposedly did.

My interest here is not value, but unresolved inconsistencies:

Why the two-font break?
Why the AP OPN instead of a more convincing AR-type identity?
Why repeat similar anomalies across multiple specimens?

Those questions remain, regardless of whether the top marking was added later or not.

[Dome]

I would also like to add one point: the discussion becoming personal is unnecessary and does not add anything to the technical topic. I assumed we were discussing this on equal footing as part of a shared hobby, focusing on the hardware and the evidence rather than personal remarks.

[xsecret]

[Dome]

Thanks for linking the article — it supports that remarked and stolen AMD parts existed. I do not dispute that.

What I dispute is turning one plausible explanation into “fact.”

Your reject-before-final-fusing scenario is technically interesting, but it is still a hypothesis about what may have happened, not direct proof of what happened to these AP-marked samples.

And there are still unresolved inconsistencies:

If these were just random counterfeit jobs, why do multiple specimens repeat similar anomalies (AP suffix, typography break, similar matrix characteristics) instead of showing more random variation?
If counterfeiters simply did “shitty work,” what is the evidence that this specifically explains these samples, rather than just being a general assumption?
And if unmarked genuine ES chips were already common, why would a profit-driven faker add suspicious markings that increase scrutiny rather than reduce it?

That logic still does not convince me.

Also, the article you cite notes defective parts would likely show instability and little overclocking headroom. Yet my sample completed long Prime validation at 2.4 GHz / 1.55 V, while my retail 3400+ CG failed that point within minutes under the same test conditions.

That does not prove a secret AMD project.

But it does make “obvious fake, case closed” sound far more absolute than the evidence supports.

Ironically, if your pre-fuse reject theory were correct, that would still make this unusual non-standard material — not a banal ordinary fake.

[Dome]

One additional point: the AP sample you posted a picture of actually appears to use a different engraving style and typography from the group of specimens I have been comparing.

That may indicate it is not even the same marking pattern I am discussing — which makes it a weak counterexample for dismissing the recurring anomalies observed in that specific group.

[xsecret]

You're free to believe whatever you want despite the evidences.

I'm done.

[Dome]

Fair enough. I appreciate the exchange, even if we interpret the evidence differently.
I am not claiming certainty, only that I think some questions remain open. For me, that is part of what makes collecting and investigating old hardware interesting.
Thanks for sharing your perspective.

[Dome]

Now, the identity is correct (and the temperatures are better too)

ibb.co/1hNJ5LX