> Intel switching to TSMC probably would have sidestepped their 10nm fab failures It would, but it would also drain a lot of investor confidence. Intel pulling

> Intel switching to TSMC probably would have sidestepped their 10nm fab failures It would, but it would also drain a lot of investor confidence. Intel pulling the TSMC card in 2024 is acceptable, or even desirable, from a shareholder perspective. Doing it in 2017 would have been jaw-dropping, even if it came with the advantages of TSMC 7nm that AMD's mobile chipsets boasted. It might have been the right call, but it would be a kick in the pants to their fab business and a terrible omen of things to come. > But, even if Intel was still doing its thing right, Apple prefers to vertically integrate 100% agree, and I feel like it's more to my point. Intel was a holdover from an era where Apple had to ditch their reliance on PowerPC. Buying pre-designed chips was an opportunistic success story on a high-margin product like the Mac, and it worked both ways until Apple started designing their own cores. Once Apple could ship an SOC that outperformed Intel, it didn't matter how much it cost to make because there was no middleman to pay. By any means necessary, undercutting Intel was a business imperative for Apple. I guess the ultimate irony of this is how I hear people begging OEMs to license ARM tech for PCs now. Intel failed as a rent-seeking middleman for high-performance core designs, so now they will be supplanted by Arm, a rent-seeking middleman for low-performance core designs. All Arm Holdings has to do now is buy a fab and they'll be a dead ringer.