Such an Interesting thread and touches on so many topic.
As I have said elsewhere in HN, and Semiwiki, I just dont see where Intel is going, not with their leadership, not on their Roadmap. It leaves so much gaps and holes to fill for a long time, delaying and changing or milking the market as much as they could without preparing for what worst might come. And this is Intel, which one of the greatest CEO of all time, also an Intel fellow Andy Grove famously said "Only the paranoid survive."
Because BK was an manufacturing guy, and in that sense not a product but business and operation guy, might be the reason why Intel has delay in manufacturing due to lacking resources. After all we are for the first time where Moore's Law, ( not in technical gate density but Economical prospective ) didn't work, and the maths and experience BK got used to for the 30 years are no longer valid. And i agree with
@raghu78 , his legacy will be for the first time ever TSMC has an arguably better node then Intel and will be shipping in larger volume with leading node. 220M+ Apple SoC every year, that is roughly the size of the whole PC market, for which majority wont be on leading node.
When BK became the CEO, it was rumored that he will continue the closer relationship with Apple. i.e Intel manufacturing SoC for Apple. Afterall he was the manufacturing guy right? But that didnt happen, nor did Intel opened up their Fab. The continue the path of trying to optimize and squeeze the process for cost, making adjustment to their tick tock to prolong product life time. And it seems didnt invest enough. 7 Years after the acquisition of Infineon, their modem are still make with older TSMC node. And god knows what the iPhone 7s / 8 will be getting, the older roadmap say it will still be with TSMC but 16nm instead of 28nm.
Seriously Infineon used to be pretty good. And Intel manage to run it to the ground.
He was rather lucky that Cloud Computing took off, Sales from DC, or basically Amazon, Microsoft, Google, OVH, Baidu, Alibaba and Tencent save his ass. But he also sat idly watching Nvidia took all of AI market share. What was Knight Milling Landing Bridge or whatever doing? Small Note; so many years later they still make shit GPU.
This Steve Jobs interview seems like it may apply to Intel these days.
Why Xerox Failed
BK is not a product guy, he is a business and sales person. The only product guy I see, Pat Gelsinger left Intel. Likely as what SJ said he was driven out of the decision making process. Pat was mentored by Andy Grove. I have always thought how Intel would be if he was the CEO and not BK. An Interesting note, IDF was Pat's idea, and it now officially closed.
GF has been doing pretty well ever since they had a new CEO Sanjay. By pretty well i mean in recent two years, so it took him 1 - 2 years to patch things up and laying the ground work. And it is all finally coming together. GF 7nm is actually on schedule for small production in 1H 2018 and volume in 2H 2018. I believe they are now facing the problem with not enough capacity. AMD Zen is selling so well, that is the reason why AMD is looking at TSMC for help with 7nm. It is highly likely the AMD Zen 2 with better IPC and 7nm wont be able to meet the demand from the market. It is likely APU, and may be GPU will move back to TSMC.
Assuming no hike up we do have a perfect storm for Intel. But of coz Intel can always lower the price, Changing roadmaps with more Cores, ( what happen to CoffeLake that was only for Desktop and not Laptop U processors ? ).
And may be, Apple A11 will be the first chip that out perform Intel's Core M at same power usage.
