Yet the core - cache aside - is significantly smaller. Supposedly on the same process but maybe maybe the transistors are built different to explain this. But

Yet the core - cache aside - is significantly smaller. Supposedly on the same process but maybe maybe the transistors are built different to explain this. But I tend to think this analysis is just ignorant. There's so many millions of things not on that spec sheet. Every stage of the pipeline has all kinds of tweakable parameters. Is the branch prediction the same? Is cache associativity the same? Are there the same number of execution units, with the same capabilities? Something changed to make the core itself smaller. We just can't read some simple specs like this & know what.