Hacker News Reader

Comment by rsp1984 | Hacker News Reader

rsp1984

Call me crazy but I think long term (10 years +) the Intel strategy of fabricating chips domestically is the correct choice. There's a ton of geopolitical pressure in this direction. The US simply doesn't want to be dependent on Taiwan for chips, for obvious reasons.

Intel knows this and is trying to become the TSMC of the west. And frankly they're the only ones that can pull it off. I'd be surprised if 10 years from now the majority of their revenue comes from selling processors rather than fabricating them for AMD, NVIDIA, etc. Why compete when there's plenty of steak for everyone?

View on HN

You may very well be right, but it's a bit much to say that's Intel's explicit strategy. In the past 12 years the main thing Intel has done is make poor long term choices: going big on mobile too late, SSDs taking off but not for Intel, whatever is gong on with 10nm.

Saying they're committed to fabricating in the US while they are, for the first time, relying on other companies to make chips for them over seas... right now their domestic fabrication may internally be viewed as a blunder.

Sure it could pan out for them but it's a crap shoot. Maybe those blunders were just all bad leadership which Intel has recovered from, only time will tell.

I am a pretty hardcore AMD fan, but fair is fair:

> whatever is gong on with 10nm

The modern 'nanometer system' is nothing more than a bizarre versioning scheme (specifically its usually the 'major' portion of a semver), and you wouldn't be able to identify a physical thing to measure to compare different fabs.

Intel wouldn't have been competitive if they were still on their own 14nm process.

Like it or not, "10nm" is the official label for Intel's most pronounced and prolonged fuck-up since the Pentium 4 dead-end. The fact that node names have little to do with node geometry is irrelevant here (nobody's trying to compare competing process nodes on the basis of nominal size); what matters is that Intel didn't finish moving their CPUs past 14nm until last year. Their recent and long-term inability to ship a functional and profitable new process across all CPU product segments combined with their inability to adjust their CPU design roadmap to account for their fab problems is highly relevant to any discussion of their future prospects, even if the discussion makes use of Intel's technically inaccurate nomenclature.

Most of the failure you're stating there predates current leadership to be fair. It's difficult to see BK as anything other than a massive failure, betting the farm on alternative growth engines whilst failing on the core mission.

Reminds me of Facebook actually - if you tell the people who build your core product that they're lame, they're goign to leave, and your core business will fail.

Spot on. Any large country that's NOT desperately building up fab facilities within the country right now is playing with fire. This stuff is the new oil. Insane how little of our modern life works without chips.

Is this about CPUs or all electronics?

Computer chips are in general a reliable hardware, and top-of-the-line CPUs are in general way more powerful than needed, so there is no need for cutting-edge manufacturing and also an embargo won't halt existing activities.

Simpler/embedded processors seem to be a generic product that can be manufactured by many countries, hindering embargo efforts.

Are chips more critical for a country than e.g. food? because many countries are not nutritionally self-sufficient.

Both - CPUs, electronics, military gear, etc.

Strategic risks such as war, invasions, sanctions, etc. have to be dealt with on a decadal basis.

You have to ask: what's the current reality, and what's the trend?

The trend is pretty clear: our world is going to depend more on computers, not less. You have to prepare for a world where drones have replaced your fighter jets, where every citizen has a smartphone, and where your financial system will collapse without working computers and servers.

> The trend is pretty clear: our world is going to depend more on computers, not less. You have to prepare for a world where drones have replaced your fighter jets, where every citizen has a smartphone, and where your financial system will collapse without working computers and servers.

That sounds a lot like the present.

CPU factories seem to cost hundreds of billions of dollars - beyond the capabilities of most nation-states. I wonder if this will cause a trend towards reducing requirements (e.g. pushing important infrastructure to rely on simpler chips) as keeping up with the technology will drain a country's resources.

I think this is completely wrong. Oil is compustible, at the end of the day you don't need a crazy supply chain to make use of it, just to make it. You couldd have 99% of a working supply chian for advanced eletronics, you're still fucked. It's far more fragile than oil.

Are you sure they will?

They seem to prefer dividend payouts rather than capex on new fabs[1]. Combined with them moving to FCF loss - I think that what you described here may be imaginary at best.

Given what they said in the last earnings call, a slow IBM-like death seems more likely.

1: https://semianalysis.com/intel-cuts-fab-buildout-by-4b-to-pa...

HN discussion on this: https://news.ycombinator.com/item?id=32291356

It's not a bad bet, if you're losing market leadership through poor execution then getting the government to prop you up as a backstop to the bogeyman seems a reasonable tactic.

Being nimble is preferred. In the long run you can vertically and horizontally intergrate every available business and infrasture. It will save cost assuming those businesses are relevant. In the short-run you burnt out before those business "expired" or fall out of trend. The trick here is to have the right integration at the moment and expand in the market direction that favors it. Is like friendster and myspace too early to boom. Is like every other zoom-like businesses too late. The west has too high labor cost and seriously shortage of engineers. Countries in Asia like China, Taiwan, and SK churned out engineers like the women studies undergraduate in the west. The business environment in Asia will simply edge out the west. Remember solyndra? Remember GE? Remember Kodak? Upcoming is Tesla. There are so many Asian side EV coming in now.

Sort of depends on whether Intel catches up with TSMC or if TSMC just extends their lead over them.

It seems pretty clear that most firms will just go where the bang/buck is (especially ones like AMD and Nvidia where performance is top priority) irrespective of geopolitics. If intel can't attract the right clients then it might make their extra fab capacity in the US economically tricky to maintain.

It's very hard to lose in the short term, year after year, and hope that a black swan event puts you back into a winning position.

Unless China's clearly stated policy that recognizes Taiwan as a part of China changes, it isn't really a black swan event.

It's more like an increasingly active volcano. It might not blow up tomorrow, but if you deliberately built a house on its edges, it would be hard to sympathize with you when it finally does go kaput.

How does one appease quarter-to-quarter returns chasing investors with that?

That assumes the only possible way the China-Taiwan issue can be resolved is some kind of event that would disrupt the production, ie. war or at least blockade of Taiwan.

Not the person you're replying to, but if I ran a chip business (or a country with a lot of military/economic tech dependent on chips-- IE most countries), it would certainly be something you'd want to de-risk if you could. Sure it's not the only possible outcome. But it's a realistic outcome, and one that has a ton of downside if it happens.

If I'm running a country or a multibillion dollar business, I'd certainly not want to leave my future wellbeing to the good sense of two powers with a gigantic historical axe to grind.

Not if you’re partially valuing your position in the stock as an option on that black swan event.

If you think there’s a X% chance of this happening, and that this event will be the catalyst for intel’s value appreciating by Y% of TSMC’s market cap, you can value the opportunity cost of investing in intel in the interim as a (dividend-paying) insurance premium covering your TSMC position.

If it never comes to pass, at least you earned a dividend. If you think Intel’s valuation has hit a trough, it’s both a good investment and a lot cheaper than a put option.

Your first para is correct but this doesn’t mean that Intel will become the TSMC of the west.

Doing so needs not only huge amounts of investment but a massive change in the business model and by all accounts culture. Neither of these are a given at all.

Given that we might be on the verge of war with China over Taiwan because Nancy wants to sip a frosty marg in Taipei, this isn't exactly sticking your neck out too far lol

Reply on news.ycombinator.com