And reactivate disabled cores? Some are just underperforming and some work fine but are disabled for demand.
>>59840860
Yes
>>59840877
Amazing,now lets talk money,what tech would I need to accomplish that? Lets say I would buy X and then offer my service to Y people,would I get filthy rich or would it be filthy failure .
>>59840860
Who made this shit? What idiot left the L2 cache on the disabled cores?
>>59840860
It depends if the Cores are locked of with microcode, or fused off at the factory. If it's microcode then yes, it happened before with Phenom.
>>59842572
That's part of the L3, not L2
>>59840860
No, they don't shut them off with firmware anymore. They laser them off.
>>59840860
If i was AMD i would simply poke a tiny physical hole in them.
>>59840860
There's been no talk of Phenom II-style core unlocking shenanigans, so they're probably lasered off.
>>59840860
if it's done in hardware with the traces? very difficult. it might not work properly even if you found a way to bridge the broken traces.
done in software? easier but still has challenges.
the issue with using binned cores is that they are usually binned because they don't work properly. they might not dissipate heat properly, they might not clock as high as the other cores, or they might not be stable for any operation and crash constantly.
>>59840911
it's not rocket science really. just looking at >>OP you can pretty much tell the obvious places you'd use to disconnect the cores
a pincette, a magnifying glass (or preferably a optical microscope to make it easier on yourself, especially if you're a bit older), thin wires and glue to connect back cut off cores.
and steady hands, of course.