"Core Temp" - A program for accurate K8 temperature monitoring

[zir_blazer]

Well, basically you can tell the actual core that is in the CPU by the CPUID and the socket the CPU is run on.
I've simply decided to go by the same route as other ID programs to prevent confusion for users.

I think that you don't get my point. The CPUID tells that my Processor is a Propus when in truth it is a Deneb, according to Stepping. The issue is that in this generation, AMD can use two or maybe more different Cores for the same model, and all got the CPUID value of the flagship Core than that model uses instead of the CPUID of what they truly are (At least hold true for the AIIX4 line, where a Propus could be replaced with a Deneb with disabled Cache L3). The only accurate way to identify them is to look at the Stepping in the Heatspreader of the Processor, because no Software application says what it really is, and that is because what AMD seems to be doing with the CPUID value.

Historically, models with different physical Cores had different CPUID/OPN even if they share the same model specifications. If there was a Core with disabled parts, it still keep the CPUID of the full version instead of being replaced. For example:

S754 Athlon 64 3000+ (2 GHz, 512 KB Cache L2)
ADA3000AEP4AR Clawhammer SH-CG (CPUID F-4-A)
ADA3000AEP4AX Newcastle DH-CG (CPUID F-C-0)

The Clawhammer had 512 KB Cache L2 disabled, and instead of mixing it with the Newcastles like if they were the same thing, they had its own OPN (AR prefix), and keep the CPUID common to all Clawhammers. Same with, for example, Manchesters and Toledos with one core disabled. However, for the 45nm K10 generation (Not sure about 65nm), this is what we got...

SAM3 Athlon II X4 620 (Quad Core, 2.6 GHz, no Cache L3)
ADX620WFK42GI Propus BL-C2 (CPUID F-5-2)
Deneb RB-C2 (CPUID F-4-2)

That is fine by itself, however, as we know, AMD may use faulty Deneb RB-C2 Cores for that model because many people could unlock the Cache L3 that on a real Propus is nonexistant. Mine is precisely one of those Denebs, according to the Stepping written on the Heatspreader. However, the CPUID is reporting the value of a Propus. I consider than that is what really is confusing people, as it would be easier if those faulty Deneb still maintain its F-4-2 CPUID for easier recognision, but AMD seems to have overwritted it to that of Propus for the Denebs that were used for AIIX4 models. Even more laughtable is that for those people that did unlock their AIIX4s, the CPUID value got "upgraded" to that of a Deneb instead of being a fixed value.

The question is, do you know for any real reason that AMD may have opted to play dirty with the CPUID value? Even if they had the same OPN, it would be preferable than each different physical core had its own CPUID (Like they did on the entire K8 era) and wasn't changed for that of another physical core if they are selling it with disabled parts.