Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Intel and the $1.5T chip industry meltdown

economist.com

Post reply on HN
Print view View original

541–550 of 734 posts

Re: Intel and the $1.5T chip industry meltdown

#541

Post by varispeed »

ansible wrote: post #263
nonrandomstring wrote: post #236

We've reached or are approaching "peak tech". The other day we were discussing Pablo Azar, "Computer Saturation and the Productivity Slowdown" (Federal Reserve Bank of New York Liberty Street Economics, October 6, 2022) [1] Simply reinventing systems and deliberately obsoleting stuff to create fake demand is over. The metaverse isn't hoing to happen. People are fatigued with tech. We're entering a different era in wh…

I disagree. We've already passed the tipping point for tech, and investment will continue to accelerate. This is true for various form of artificial intelligence, and the rest of tech. Businesses in general started to see the appeal around the time that the Google Assistant, Cortana, Alexa and Siri started to become actually useful and desired by the general public. Because of that, there will continue to be a push t…

> The current systems can turn a normally worded question into a web search, and present the results in a pleasing manner.

Pleasing but not useful. I really miss Google search from 10 years ago. I think things started going downhill when they introduced the instant search feature. Nowadays it's difficult to find anything. The results are mostly irrelevant and littered with spam and AI written nonsense. I remember I could go through hundreds of result pages and find useful things on the tail end, now quite often I get one page only and nothing is relevant to my search term. I have to often use quotation marks etc, but even that stops being helpful - basically more often than not I get no results at all.

Re: Intel and the $1.5T chip industry meltdown

#542

Post by Rury »

PuppyTailWags wrote: post #314
ansible wrote: post #263

Earlier quoted context omitted.

I disagree. We've already passed the tipping point for tech, and investment will continue to accelerate. This is true for various form of artificial intelligence, and the rest of tech. Businesses in general started to see the appeal around the time that the Google Assistant, Cortana, Alexa and Siri started to become actually useful and desired by the general public. Because of that, there will continue to be a push t…

I disagree current systems can turn a web search that presents results in a pleasing manner. I had a horrible time trying to find a specific musician's work on YouTube just last night. It kept recommending me completely unrelated vids as if I was browsing just to browse. It couldn't seem to comprehend that even though I usually watch cute cat videos today I am searching for music, even when I was exact quoting the au…

I think a part of this problem is just simply a matter of scale. As more and more information gets generated and stored on the internet, finding the information you exactly need in 3-5 words typed into a search engine at some point just ceases to become probable or even possible. So search engines have to predict what you want based upon the limited information you give it (and does this using others things such as how often something is visited, search patterns, trends, and/or related promotions), and you end up perceiving it as it telling you what you want.

Sadly giving a search engine too many words can stop it from even searching, probably because that would cause a scan or result in a very time and computer/resource intensive search. So it doesn't even bother searching and tells you it couldn't find anything.

Re: Intel and the $1.5T chip industry meltdown

#543

Post by dirtyid »

greycol wrote: post #478
dirtyid wrote: post #438

Earlier quoted context omitted.

Depends on how Sino-US competition veers. As important as TW is to rejuvenation narrative, it's ultimately the consolation prize versus dismantling US east asian security architecture and securing regional+ hegemony. That's the grand finale battle for the lightcone of future PRC security/prosperity. >Japan is very able to defend Japan (and SKR, and TW, and even PH) like most US allies in island chain are are heavily…

The issue is in such a war China is also cut off. A major point of their south china sea claims is to ensure that there is no peaceful way to block oil (and other products) being delivered to Chinese ports. In an attack on Japan (that doesn't turn into MAD) China also loses this supply chain and becomes reliant on Russia for energy imports and the infratructure for that reliance isn't in place yet and is also a major…

PRC is essentially calorically food secure (with huge waste / room to optimize), has large energy reserves, and unlike island nations, massive domestic raw resource supplies. PRC is NOT Japan during WW2. Hence PRC is much more autarkic and can drag on war economy, perhaps indefinitely. Sure people will eat less meat and depend more on EVs (maybe even cope on bikes) during transition, but when shit hits fan, PRC + RU (is a powerful self sufficient land bloc with much greater long term war making potential than US partners trapped on vunerable islands. It's about asymmetric vunerability.

>The issue is in such a war China is also cut off

The Malacca dilemma was based on assumption that US had unilateral power to blockade PRC imports with impunity due to being domestically energy secure - it was an argument/strategy also based on asymmetric vunerability.

But that's increasingly not true, the TLDR is PRC rocket force likely already has capability or will in short term to _conventionally_ strike major US energy infra... US is existentially dependant on ~150 refineries - they are as dependant on these refineries as PRC is on maritime energy shipping. People conflate resource security as having more resources in your soil but it's really about the ability to protect the critical extraction/delivery infra. Otherwise Saudi wouldn't bribe US for security. Obviously conventional CONUS strikes is also a prelude to MAD, but it is also an equation for PRC establishing mutual vunerability with US, which greatly constrains US actions. Not to mention such capability also functionally dismantles US naval supremacy via port strikes (both capital and support assets) that underpins US global power projection.

My feeling is that the chance of US blockading PRC when she becomes as (conventionally) vunerable as PRC is increasingly remote. It's hard to understate how much geostrategic calculations must change once a relatively autarkic industrial power as massive as PRC is able credibly bring actual war to US homefront. It will be first time in modern history where conventional fires can penetrate CONUS to meaningfully degrade US society. US will have to assess whether it wants to fight a possibly existential war (possibly at best a pyrrhic one where she might not uphold her hegemony after) or abandon East Asia where PRC preponderance is increasingly difficult to match or deter, especially with respect to TW.

Re: Intel and the $1.5T chip industry meltdown

#544

Post by CydeWeys »

terminalcommand wrote: post #287
kuschku wrote: post #85

Earlier quoted context omitted.

You need certain machines for EUV. There's only one manufacturer (ASML) for those, and they're at capacity. And even if they wanted to scale up, there's only one manufacturer (Zeiss) for the EUV mirrors required for those machines, and all their capacity for the next years is already bought as well.

Why can't ASML scale up? I don't understand this. I think because they are a monopoly they can work slower, they don't need the extra profits.

Why didn't they simply build more chips in 2020 instead of shuttering all of those automobile factories?

At some point you're running into actual, real constraints, real bottlenecks in the supply chain, that will take years to resolve. You can't just scale up on short timeframes, no more than you can make a baby in a shorter time frame.

Re: Intel and the $1.5T chip industry meltdown

#545

Post by CydeWeys »

chinabot wrote: post #521
nonrandomstring wrote: post #236

We've reached or are approaching "peak tech". The other day we were discussing Pablo Azar, "Computer Saturation and the Productivity Slowdown" (Federal Reserve Bank of New York Liberty Street Economics, October 6, 2022) [1] Simply reinventing systems and deliberately obsoleting stuff to create fake demand is over. The metaverse isn't hoing to happen. People are fatigued with tech. We're entering a different era in wh…

My laptop is a 1000x the power of the one I had 15 years ago yet I am still stuck with the same hourglass due to shit, bloated and over-complex software stacks. Tell me I'm at peak tech when this is no longer the case.

There's no way this is true. You simply aren't accurately remembering the actual experience of using that laptop 15 years ago. For starters, the laptop 15 years ago was using a spinning rust hard disk, so everything was much slower compared to today's lightning fast SSD. I remember when it used to take minutes to boot up a laptop and launch a few applications, and it took almost that long to resume from hibernation. Modern laptops are much, much faster.

Re: Intel and the $1.5T chip industry meltdown

#546

Post by CydeWeys »

fnordpiglet wrote: post #361
Melatonic wrote: post #355

Seems like a dumb article - their whole argument is that we should NOT be building fabs in the US because demand has recently fallen and we MIGHT end up with too much supply? Seems like a good thing to be honest. More supply means cheaper prices and even if demand is a bit lower there is no way it is going to keep going down. We have CPU's in damn near everything these days. The drop is probably mainly because of cry…

Seems like a good thing as a consumer. As a producer it’s what you try to avoid. They’re not charities

It's reasons like this why the DPA and other mechanisms exist. Having sufficient fab capacity in case global supply lines experience disruption is a strategic reserve of high national security importance.

Re: Intel and the $1.5T chip industry meltdown

#547

Post by tuatoru »

xyzzyz wrote: post #522
iepathos wrote: post #515

Earlier quoted context omitted.

The western world would not be compelled to engage militarily over Taiwan. We'd only need to provide money and weapons to Taiwan (rank #21 military in the world) to hold off China (#3) indefinitely given the fact that Taiwan is incredibly difficult to invade due to the barrier the mountain ridges form around it and the narrow straight leading to it for a sea attack. It's a natural fortress. Look at the ongoing failur…

How do you provide weapons to Taiwan if Chinese Navy blockades Taiwan?

Ahead of time.

Re: Intel and the $1.5T chip industry meltdown

#548

Post by godelski »

Melatonic wrote: post #355

Seems like a dumb article - their whole argument is that we should NOT be building fabs in the US because demand has recently fallen and we MIGHT end up with too much supply? Seems like a good thing to be honest. More supply means cheaper prices and even if demand is a bit lower there is no way it is going to keep going down. We have CPU's in damn near everything these days. The drop is probably mainly because of cry…

Oh no, if only it were possible to expand the use case of cheap and affordable chips. It's not like we can put computer chips into devices like refrigerators, thermostats, watches, lights, tea kettles, ovens, cars, bicycles, musical instruments, desks, keyboards, doors, sorry, I lost track what we were talking about. Seems like we can put computers in literally everything. Isn't that what the IOT people promised us?

Re: Intel and the $1.5T chip industry meltdown

#549

Post by godelski »

theptip wrote: post #288
LeifCarrotson wrote: post #204

Earlier quoted context omitted.

A fab in Ohio, Oregon, or Arizona 10,000 km from China is safe from threats like cruise missiles that the proposed fab 1000 km from China in Japan or South Korea would not be. It's not so much that cruise missiles used on civilian targets in South Korea or Japan would not be a reason to start WW3, more that it would be less tempting.

I'm confused by this take. Who is threatening Japan with cruise missiles? Certainly not China or Russia. North Korea, perhaps. The primary reason that we're concerned about semiconductor fab concentration in Taiwan is that China has consistently stated that it is going to invade Taiwan at some point (and that could be 2049 or in a few years for all we know). This is completely outside of any hypothetical scenarios of…

> I'm confused by this take. Who is threatening Japan with cruise missiles? Certainly not China or Russia. North Korea, perhaps.

Countries that have actively threatened Japan with nuclear bombs: China, Russia, North Korea, and the US (which literally dropped a bomb but have yet to make threats since).

Re: Intel and the $1.5T chip industry meltdown

#550

Post by agumonkey »

yazaddaruvala wrote: post #372
agumonkey wrote: post #369

Earlier quoted context omitted.

In terms of human existence tech is overwhelmingly above the average person need. It's not providing anything beside more addictive paths. You don't need more resolution, more bandwidth, more anything.. you probably need less, or more space to reflect and live outside the networked digital realm. ps: to add a bit more, from the few chats I have, people are either saying "well I need that to check my bank account or p…

> You don't need more resolution, more bandwidth, more anything.. you probably need less, or more space to reflect and live outside the networked digital realm. Better tech is how humans (all of them - rather than the few) get more space to reflect and live outside.

That's shallow. Current tech does not help. You have to filter data, organize, manage bookmarks, fomo, notifications, slippery software.

Better is subtle. Nothing is a straight line.

Post reply on HN