Let's see. Up to now, all I have seen is marketing gimmick. Intel's new management just makes new ads with "select" benchmarks to show that its processors are better in some tasks than Apple's M1. Now Intel is renaming its processes because they do not reflect reality. Too many words so far, but zero in actual performance improvement. The chips will not become faster just because Intel renames them.
But that's all they have. Seriously.
I remember being an IPD (Intel Products Dealer, used to be Intel Processor Dealer) and they were going on about how awesome it was to have so many cores, and how awesome it was to have so many cores with multi-processors, and then people from the audience started picking them clean.
Why do some benchmarks get worse with multi-core? Why do some benchmarks get worse with multi-processors? Why do some situations work blazing fast with smaller amounts of memory, and dog down on larger, and sometimes the other way around? And into servers too. I ran into a server with 2 Xeons that would rock the world with x-amount of memory, but plus or minus the next options, the whole thing performed so much slower. But Intel wasn't about to get into why their stuff wasn't great, they would taught their own weird benchmarks showing their stuff rocked, and use weird inconsequential ones to rip their competitors. It was like high school all over again...
It's all just crap marketing. Like when they had the first multiprocessor systems, if you used the second socket, some things to a HUGE hit. So they would say 'Sell the second processor first', and then their own boards would dog after adding the second processor, and they had no clear answer why some tasks were faster, and others were just completely underground.
But that's 'marketing'. Polishing those turds...