Intel had no real intention to stick with TSMC though, so once again everything hinges on Intel 18A. And reading between the lines, Intel Foundry doesn't expect many customers until afterwards - so Intel 16A etc.I imagine Intel will reach parity with AMD with their next gen after Arrow Lake and then probably succeed them - as now the process gap has been eliminated.
Therefore even if Intel and Pat's hype about returning to process leadership are true - and given Intel's history of hyping process since at least Intel 10nm their credibility is not great - the $10's of billions for the 18A fabs have be paid largely for by Intel's products alone.
How Intel's huge P cores will fare - Skymont is nearly a third of the size, Zen 5 is smaller on the lesser N4P process (and includes AVX523 which used to be Intel's great hope), etc. - is not certain. While Skymont is impressive, mainstream requires an ST champ and that it is not.
The proper Alder Lake reviews with deep dives (maybe chips & cheese) should still be interesting.
Last edited: