I found a very long and insightful article on SemiAnalysis today on how Intel has been doing the wrong thing for so long, it’s really hard for even the people vocally supporting them to see that Intel will regain their once-beheld throne.
They were supposed to stand up fab capacity in North America funded in part by the chips act ($8.5B), clearly this did not happen. Some $500M was pulled back by the awarding body, and. TMSC have established the first of a number of sites to support local customers across the USA.
TSMC sees early success at first U.S. chip plant.
2 Likes
Their leadership grossly underestimated and/or lacked appreciation for the challenges that were ahead of them back at 14nm.
I was alarmed when Intel was under pressure at 10nm and decided they’d change their process naming from traditional transistor size to some arbitrary marketing code:
10nm is Intel 7
…
5nm is Intel 4
And then realignment at 2nm
2nm intel 20a
1.8nm intel 18a
It was as though they were trying to hold off market until they could catch up, but that will never happen now as the gap is widening and the leaders (Samsung/Tsmc) are pulling away.
Intel as it is, failed
Idm2.0 had no chance as the investments of both architecture and fabrication are far too far for one company to excel at both.
The fault falls squarely on top leadership and the actions were just, if not tardy.
Fracturing the company is essential for the western world to remain relevant.
Restrictions against fabrication suppliers (LAM/TOKYOEL/ASML) will break down if there is nobody to sell to (ie no demand from Intel), otherwise innovation would be stifled and forward progress would be impeded.
Humanity is racing against time, and transistors are our only apparent solution.
1 Like
The ARM CEO comments in this interview that he offered to use Intels fab capacity to keep the pipes smoking. PG declined
1 Like
That’s just wild. I could see that as a negative from a market view perspective, but it would be money in the bank.
1 Like