Seems like TSMC's days are number, so... is the entire western world just gonna accept a Samsung Foundry monopoly? Feels like Intel is Too Big to Fail (even if they keep falling behind their own timetables).
Also it feels like a make-or-break moment for a lot of things like Gaudi, Xe, and the chiplet focus.
I don’t know if TSMC’s days are numbered, but at best they seem to be fumbling with their 3nm process and it kinda seems like Intel is hitting bottom as far as their process development woes are concerned. Theres an intel 7 refresh up next rumored at the end of the year for consumers, and then they will seemingly take their next swing at gpus with Battlemage and should be ready to release intel 4 consumer products next year. Imho, at least, they seem to have the most exciting stuff coming next year.
I don't trust any single bit of "it's coming" from Intel after the 7 years delay of their 10nm node and their disappointing 7 one. 7 years is a disaster and a dead sentence in this business.
They aren't the new IBM/Oracle yet, but they need a bet that doesn't depend on their manufacturing capabilities.
The only refresh I heard of is the Raptor Lake CPU refresh which will be Intel’s 14th gen CPUs - which might enable some features like the Digital Linear Voltage Regulator (DLVR) that were disabled due to bugs in the 13th gen CPUs; DLVR could potentially significantly reduce power draw.
The US government will backstop intel if need be. Losing the ability/access to manufacture high end silicon is near the top of national security concerns. We can't even do the cutting edge right now, and that's why the chips act was passed.
generally agree but the anecdotal evidence I heard from a former employee about the internal dysfunction has scared me away. apparently everyone in a leadership position has a rudimentary understanding of the products they build and is chasing off all the talent like Jim Keller who are forced to deliver the bad news that they’ve made a mess they cannot easily clean up without abandoning current initiatives and starting anew with more knowledgeable leadership
But even if Taiwan isn't invaded, TSMC could still be compromised. For instance, they could suffer from a severe peacetime cyberattack, physical sabotage, a blockade of the island, or even get damaged in a failed invasion attempt or a brief exchange of fire. They are a high profile target.
I am not that gung ho/confident about U.S. protection, and I'm not even sure China is that aggressive, but it doesn't really matter because the liklihood of TSMC being damaged through subversive action is much higher. China very explicity wants to coerce Taiwan before "resorting" to invasion, and damaging their international trade is one step towards that goal.
The US would drag all of the rest of the civilised world to war with it, as they kinda did with the proxy war of Ukraine. The US+allies could probably take down China+allies. But if this happens, we are all, gloriously fucked.
I wish countries in macro level geopolitics did not behave like prepubescent boys trying to share the same toy
Not the OP, but I imagine there's at least some market watchers who assume the PRC will gobble up Taiwan sooner rather than later. Intel seems like a poor alternative, but if it's all you had...
A lot of thing, China's economy and the country itself, might be numbered if they decide to invade Taiwan.
This isn't like Ukraine where there was an implicit agreement that nobody will send actual troops into Ukraine to fight its war. If Taiwan is invaded, the US will send its own troops. And if the US is involved, the EU is likely involved too. China is big but not enough to challenge the US militarily. Not yet.
It would take an absolute madman to make China go to war just for Taiwan.
The problem with China is that it's been building its army up steadily for quite a while, and it's quite large, but also probably expensive.
And, if you've got a big army (and navy), just sitting around doing nothing.... what are you going to do with it?
The army themselves must be itching for a fight, just to justify their existence. Maybe that's why they're trying to hot things up in Bhutan.
It’s in nobody’s interest to start a war over this. It’s all fluff.
There’s a cottage industry of war hawks who sell various narratives about China to rile up the mostly republican base.
My favorite is Gordon Chang, who a few years ago was predicting that the Chinese communist party and state was going to collapse and start some sort of war. Now the Chinese are beating the USA in space and are about to invade Taiwan.
It’s all stupid. Semiconductors are a high risk capital intensive business. Just like Verizon bet and lost on mmWave 5G and is getting its butt kicked by TMobile, Intel bet on 450mm wafers and lost.
Strategically, they’re in a recoverable position. TSMC is at capacity and needs to invest billions and execute to maintain their lead. Cloud is challenging their margins, but their relationship with Microsoft gives them leverage.
People said this about Russia for the longest time and were wrong. Dmitri Alperovitch predicted the Russian Invasion 3 months before it happened and he’s predicting a hearty chance that China will attempt to invade Taiwan in 2025 - 2030. If you don’t have similar credo about past events like that, then I wouldn’t trust your opinion over his.
Russia isn’t China. People may have disagreed over the timing, but the Russians already seized crimea and interfered with US elections to advance their Ukraine policy.
Russia is a declining petrostate. China is the world’s factory. Antagonizing their biggest trading partners won’t be a winning strategy for many years, if ever.
Gordon Chang is a favorite straw man for Chinese nationalists. They basically just mention him as a summation of all western reporting on China.
My favorite is Andrew Left, who shouted the alarm on EverGrande in 2016 and then was fined for it in Hong along because he obviously didn’t understand the Chinese market.
> Now the Chinese are beating the USA in space and are about to invade Taiwan.
China isn’t beating the USA in space by even the most pro Chinese nationalist measurements.
We're only just starting to see Arc mature, and their line is looking incredibly impressive given their success with optimizing Cyberpunk.
On the CPU front though, it seems they're continuing to lag. I suspect they may have over-invested into GPU's when mining was booming, and the risk isn't paying off.
They'll be back, but I don't see them being a major competitor for many more years outside of commodity and Enterprise business. Which, as a company, is a great place to be. But I really don't care.
Aren’t 12th and 13th gen Intel CPUs going toe to toe with Ryzen 7000s and sometimes with more cores/$ (albeit some efficiency cores)? Thats what I see anyway, I heard AMD is winning over the server market but don’t follow too closely there.
Yes, but the problem with Intel CPUs is that they have to pull significantly more power than AMD does to get that performance- that's kind of why they have E cores and have to provide more P cores per dollar by comparison.
This is why I don't invest in single stocks. It was clear back in 2015/16 that Intel was in quite a bit of trouble but their stock continued a really good run up. Even the board was clearly worried by 2018 and got rid of the CEO on a pretext. I've no real information of how they're doing internally now but they've clearly got some good people and put some better leadership in, but now their stock is languishing... I guess they've just got to white knuckle it through the next 5 years and hope they deliver.
Absolutely no evidence to back this up, but if I were the Taiwanese government, I’d be making sure TSMC was slooow-walking their overseas expansion plans.
Seems like TSMC's days are number, so... is the entire western world just gonna accept a Samsung Foundry monopoly? Feels like Intel is Too Big to Fail (even if they keep falling behind their own timetables).
Also it feels like a make-or-break moment for a lot of things like Gaudi, Xe, and the chiplet focus.