TSMC isn't a monopoly. You could go to Samsung like Tesla did. People weren't bringing monopoly worries in 2018 when Intel shut down their first disastrous attempt at being a foundry after about 5 years.
What you're seeing now is an "offer you can't refuse" to the fabless designers to direct private capital to Intel by the US for national security reasons and some political points. It buys Intel cash which it desperately needs for its foundry efforts and is a quasi-bailout of Intel shareholders and debtholders. The problem is that Intel doesn't know how to be a foundry in terms of volume and breadth of customers and will need an indefinite amount of capital to learn.
> The problem is that Intel doesn't know how to be a foundry in terms of volume and breadth of customers and will need an indefinite amount of capital to learn.
Indefinite is much less than infinite, they will learn because they have to. For better or worse, throwing money at the problem is how the US operates, trowing trillions into AI cannot risk being dependent on unstable Taiwan or capacity-constrained Samsung.
It will be more than just Apple. Trump got in first. He'll arm twist the others to take token stakes and set up some working relationship. Nvidia doesn't need to take a $5B stake in Intel. Did Apple ever have a stake in TSMC? Whether or not this changes any of Intel's actual problems over the long term is a different matter.
I talked about this in another comment here, but the danger in unhealthy introspection is getting lost in your own mind in an Escher stair fashion where the immediate scope of your thinking appears to make sense. But in reality you've constructed a faulty, hard to escape representation.
It takes some time, but it helps to take a somewhat adversarial approach and try to falsify your introspection. That's why writing is important. It forces you to lay out the structure for later review. When you just think about something, how that thinking "felt" makes up a lot more of your judgement than the actual coherence or validity.
You need to be able to step outside yourself which is easier said than done, but that's a good skill for a variety of reasons. Writing is an important way of doing it. It's like a snapshot of you in a moment of time that you can review later when you are a slightly different person. Sharing your introspection with others who you trust but don't think like you can also help: good friends, therapists, mentors. LLMs can be great for this sort of thing if you approach it with healthy skepticism and avoid leading it to your answer of choice (but if people can't, who knows where they'll end up).
Setting time boundaries help a lot. Also, resetting your mind by doing something that doesn't bend much to navel gazing helps. Your garden doesn't give a shit about your thoughts. I took judo on a trial run once. Your mind clears very fast when somebody is trying to re-introduce you to the ground.
I think that today, people's ability to do this has greatly diminished. Technology has made it much easier for people to get trapped in those staircases just because it makes them feel good in the short-term. With networks, it's even worse because now they are comforted by the shared experience. Those staircases become the identity and reality rather than something to escape from.
I journal to create a concrete representation of my thoughts at a point in time. Writing with intent to explain to yourself forces a structured view on assumptions, deductions, inductions, etc. rather than having them bounce around in your head as a wobbly mass of jello.
The structuring of your thoughts and emotions, by nature, forces a certain amount of meta cognition and distance from your feelings as you can more easily identify and further think about the insights, inconsistencies, etc. You have a scaffolding that you can continue to build on or rebuild.
Beyond the immediate thinking, it's also interesting to see how the thoughts on something change over time or how they relate to other topics. Now, I'm thinking about my thinking rather than just focusing on the immediate topic.
Deep thinking without a sense of permanence can be great as a start of ideas. You need a low cost way for ideas to form where you can pick the most promising ones for more introspection. But without permanence, there's this danger that you get lost in your half-formed thoughts because it has a lot of "feel" in it and makes it hard to build on or critique. There's a danger that you end up going down an Escher staircase mentally and emotionally that seems like the logical, immediate thing to do but is a contradiction at a higher level.
One CEO is gone mainly because he thought that he could home run his way back to prominence with a "Fabs to Nowhere" strategy which has crippled the company's finances.
One "Product CEO" had a material chunk of her responsibilities re-allocated to Tan after he came on board with him frequently saying that their product roadmap has been subpar.
I'm with you. For some reason, the hardware / tech crowd insist that evil finance people or evil MBAs killed Intel. It's such lazy thinking.
But if you look at the past CEOs, Intel had Krzanich (fab guy), Swan (CFO guy who didn't even want the job but they couldn't find anyone else), and Gelsinger (design guy and Grove disciple) in 11 years. I'll even throw in Jim Keller, not a CEO but still The Chip God, who left in frustration after two years.
What's the one common problem that all of them had despite all their different backgrounds? Getting relevant nodes to market and scaling them up. Their foundry efforts (v1 and v2) have been disastrous. The CEOs or MBAs perhaps were a friction, but they aren't the root cause. Technology Development has been the center of power that the fabs and products revolved around for decades.
I blame Bill and Sohail, principally. I think they essentially orchestrated a coverup of TD’s growing problems in the era of PSO and BK should have known better but was out to lunch.
That's what the market has been saying for a while: the cost of the assets were mal-invested in its current context and are not worth what the accounting says they are.
Markets obviously can be wrong, but to think that markets are so inefficient that it confuses book value with liquidation value for a company like Intel is much more wrong.
Intel has been backpedaling away from the idea of 18A as an node for external customers for a few quarters now as the external demand for 18A is very low. CFO has said external customer revenue isn't expected to have a material financial impact even by 2027. Intel has been pushing the narrative that 18A is primarily an internal node from the start.
I think what's more disconcerting if you are an Intel fan is that Intel likely now recognizes that 18A will not have the impact that they originally thought when the ex-CEO "bet the whole company on 18A." Rather than dump more resources into it, they are instead trying to get ahead of the problem by moving resources to 14A which in theory has broader appeal.
Work is an intersection between what the company wants and what I'm willing to do. I can leave the company, try to change it from the inside and accept the consequences, or adapt to the situation until I can find or create something better.
But I find that whinging that work should fit my worldview becomes dis-empowering quickly. The temporary dopamine hit from getting sympathy venting and believing that the rest of the world is crazy makes me feel like something has changed when nothing has or perhaps will. And that just causes me to feel worse over time.
It's also interesting to see how opinions of what is fair, right, reasonsable, etc start to change as the contributor becomes a manager of a team and then an owner. The individual's context can have a lot to do with shaping the opinion.
Intel has practically nothing to show for an AI capex boom for the ages. I suspect that Intel is talking about it early for a shred of AI relevance.