Neon

Joined: 04 Feb 2008 Posts: 1512 Location: Dallas, Texas, USA
|
Posted: Sun Sep 12, 2010 3:02 pm Post subject: Early Athlon 64 - BIOS disagrees with some part numbers |
|
|
I obtained an early Athlon 64 ADA3400AEP5AO with date code 0326, well before the Athlon 64 product launch during week 0339.
The BIOS reports it is AMD Athlon 64 3200+, not 3400+ - I was surprised.
Is it a fake? The font looks OK, the package number 27872 is correct. There is no evidence that the heatspreader has been removed and replaced with a different model heatspreader, such as scrape marks or broken sealant. If you think it is a fake, let me know why.
However, there is another early Athlon 64, ADAAA2800ACN5. The BIOS identifies that processor as 2600+
It seems that in first half of 2003, AMD had not yet decided upon final performance ratings for Athlon 64. So I compared Athlon 64 socket 754 processors with 1024 KB L2 cache.
In the table, I split Athlon 64 into early parts on the left, and later parts on the right. It appears that AMD initially used performance ratings that were too high. Two examples: a) early 1800 MHz parts received 3100+ model number, but in later parts this was revised down to 3000+; b) early 2000 MHz parts received 3400+ model number, in later parts revised down to 3200+.
I believe this is why current BIOS report the ADAAA2800ACN5 as 2600+ model and the ADA3400AEP5AO as 3200+ model - these model numbers more accurately portray actual performance than the initial model numbers.
If anyone has the ADA3100AEP5AJ or ADA3100AEP5AO parts, I would be interested to know if the BIOS reports these as Athlon 64 3000+. |
|