Core Color vs. Stepping

I have two XP 2500s, here are the stepping infos with the differences highlighted red:

CPU A
AXDA2500DKV4D 9886358271163
AQXEA 0324 TPFW
27488 E5 T <-- this being the stuff written along one edge of the CPU

CPU B
AXDA2500DKV4D 9886358271161
AQXEA 0324 TPFW
27488 B1 T

They should be identical from 'standard' stepping comparision -- both are DKV4D AQXEA 0324 TPFW and differ only by 2 on the long number (they were produced 2 CPUs apart?) -- but CPU A has a bluish core while CPU B has a rose-colored core. Unfortunately I haven't yet been able to test the processors head-to-head for overclockability.

So I'm curious:
1. What causes core color?
2. Why would very identical steppings have different core colors?
3. It's generally thought that CPUs of the same stepping (especially sequential steppings, like these) came off the same (or same batch of) wafer. Does this suggest otherwise?
4. Could this have overclocking implications even though the stepping says that there shouldn't be a performance discrepancy?

Well some numbers and/or letter have to be different, or they would be the same CPU (code wise), right? AFAIK there aren't any Barton CPUs that have exactly the same data on them (someone correct me if I'm off.

And, I don't think there are any overclocking implications in the stepping differences.

test em