- Most aptly described by CB's as a "creosote bush" x86 and its high margins poisoned the ground. There were multiple attempts to diversify the product line from the mostly stupid toys, to the high end graphics chips, to the "me too" efforts to recapture phones, to acquisitions but the moment they failed to track toward x86 level margins investment and resources dried up and guaranteed their failure.
- BH's ill-fated decision to drive cost per transistor for 10nm along Moore's Law. While T and S were allowing cost to transistor to increase, B's decision committed Intel to radical degree of backend scaling. Had Intel been able to pull it off it's technology leadership would be today unquestioned and unthreatened. But it proved a reach too far exacerbated by
- An obsession to protect intellectual property. In this the module engineers were allowed to know only the specifics of their process steps. This placed the entirety of problem solving on the shoulders of a small number of integrators who knew about a specific process segment, and the one who knew the entire front half flow, the one who knew the entire backend (interconnect) flow, and the one who had access to the entire process flow. By "design" the number of people capable of solving problems was reduced as the number of problems created by #2 exploded and the development machine was self-strangled.
An on point post that needed its own thread. I hope @1wkq+1iOHER9S doesn't mind the repost.