Relationship of L2 cache and overclockability in 90nm SOI Sempron

[sangyup81]

This is a continuation of this thread.

Here are the 128k L2 cache versions:
2600+ 1.6ghz $79.99 Newegg shipped retail
3000+ 1.8ghz $99.00 Newegg shipped retail

And here are the 256k L2 cache versions:
2800+ 1.6ghz $89.99 Newegg shipped retail
3100+ 1.8ghz $116.00 Newegg shipped retail

My theory is only based on 3 people, Me, Avalon, and Budman and a review from XBit Labs

Here are the respective OCs
Me: 2600+ 128k 1.55v 306x8 = 2448mhz (DFI)
XBit Labs: 2600+ 128k 1.65v 313x8 = 2504mhz (DFI)

Budman: 3000+ 128k 1.6v 300x8 = 2400mhz (Chaintech and DFI)

Avalon: 2800+ 256k 1.55v 340x8 = 2720mhz (DFI)

And recall the old 130nm 3100+ were 256k cache ones and could OC to the high 2500s. Could it be that the 128k cache chips have defects that the 256k cache chips don't?

 

[Avalon]

Actually, I overclocked to 2.72ghz on my chip with 1.55v, but I only wanted to run my chip with .1v over stock, max. It was prime stable at those speeds, though. Not worth the increase in heat to me.

*Edit*
If someone will buy my chip off me for $90, I'll gladly purchase a 2600+ or 3000+ and see how they fare in my system