Hacker News new | past | comments | ask | show | jobs | submit login
Samsung plans $17B chip plant in Taylor, Texas (datacenterdynamics.com)
586 points by kungfudoi 6 days ago | hide | past | favorite | 425 comments





After the TX grid failure last winter, it's probably not a coincidence that Samsung's new facility will be near ERCOT's operation center in Taylor, which manages the TX grid, and will likely be the last load to shed when the grid's stressed. Used to work in the Austin fab, and the amount of money lost per minute in a power failure is mind boggling. The tax breaks Taylor offered ($314m) are not that different from what Samsung was reported to have lost due to the grid failure ($270m).

The ERCOT campus has their own generation capability along with sitting between two distribution grids. Same with their DR site in Bastrop. In the case of a black start condition (grid totally down), ERCOT would need to come up first to coordinate the grid restoration.

Taylor is an ideal place for a large manufacturing operation since it's close to major highways and railroads, not too far from Austin, and land is very cheap.


> land is very cheap

*was

I'm between Taylor and Austin (Bagdad Rd, Leander). 1 acre lots along this strip have gone from ~$115k a year ago, to some recent sales at ~325k.


This...but did Samsung quietly negotiate a "any further energy supply f*ck-ups are 100% on YOUR dime" deal with the State of Texas, or what? By several accounts, the TX grid got darn close to collapse last winter - at which point "ERCOT Op Center is next door, and we get top priority" would not get you a single erg.

The going deal is, if there’s a power failure, we’ll let you recoup your costs from the taxpayers even as you profiteer. The governor can be bought for a very affordable $1 million. Certainly the incentives are aligned to solve this problem! https://www.texasobserver.org/after-kelcy-warrens-energy-tra...

If the costs do get passed to the taxpayer, let's hope the populace elects representatives who openly rip up the old contract and say "we're not paying". Flex the fickle power of democracy on the investors.

Also. Property tax in Texas is high relative to other states. So it makes sense to give Samsung a break on Ad Valorum taxes (not that a corporation with sufficient lawyers would pay them anyway.) They'll just make up the difference by stiffing it to the owners of all the new houses near the plant.

I can't imagine Samsung would put a plant in a 3rd world country like Texas unless they were getting significant tax deferrals from the state.


It remains to be seen if shifting the cost to taxpayers is harmful to the economy or not.

The ONLY reason to care about the "economy" is the secondary benefit it provides the "taxpayers". The health of economy and the people should never be weighed against each other.

Sure, lets analyze who benefits. The US benefits from national security by having a semiconductor plant in the US. The local economy benefits in Taylor Texas. Companies benefit from having a close semiconductor plant. People benefit from low cost chips. I think you may be advocating for more welfare analysis, which I completely agree with.

...but it will almost certainly hurt the taxpayers as a whole.

Not necessarily. Let's review our options: regulate companies to account for every disaster scenario, let the government plan and provide assistance during disaster scenarios, or do nothing and let people suffer and let companies go bankrupt. Passing costs to the taxpayers isn't the worst option.

Bad logic. If the governance e.g., TA is not manipulating market e.g. giving this sort of subsidies in case of failure then market could find the right place. Instead of behind the scene talks, a measurable study would be done and then fabs would be finding a suitable place to establish. I am not saying Samsung has not done studies on the location. Just pointing out the fallacy in the dangerous logic.

That's a lot of what-ifs with the assumption of corruption. There is no fundamental reason why companies and the government can't cooperate for mutual benefit.

There’s plenty of precedent for corruption with respect to tax incentives, see, e.g., what Foxconn did in Wisconsin:

https://www.theverge.com/21507966/foxconn-empty-factories-wi...

https://en.wikipedia.org/wiki/Foxconn_in_Wisconsin?wprov=sft...


I'm sorry, am I missing some sort of evidence of corruption for the Samsung plant? Is that a serious accusation?

Evidence of corruption in Samsung itself is not hard to find. Absence of evidence of corruption in this specific case is not evidence of corruption’s absence. I remain skeptical of the terms of the deal either way.

Evidence of corruption in ERCOT is not hard to find.

It would be much harder to find evidence of fair dealing from ERCOT.


even if it's corrupt, although it's rare, sometimes corruption does serve the public interest, even if by total accident of competence.

Corruption as a least-bad alternative is not much of an endorsement, and likely isn’t a good investment of public expenditures.

Letting companies go bankrupt is a good thing. Letting people suffer without power is not.

Crony capitalism, the best kind of capitalism eh?

Hacker news comment of the year

> erg

For anyone wondering, this is a unit of energy (100 nJ):

* https://en.wikipedia.org/wiki/Erg


(For those wondering, it's a unit from the old cgs system, rather than the mks we typically use today.)

FWIW, the problem wasn't a generalized grid management problem. It was a failure to properly winterize gas plants for rare freezing events. The Texas grid itself is in reasonable shape as far as I know.

Also, as I understand it, there's another "grid" which failed, which is the network of natural gas wells and pipelines. Natural gas wasn't being pulled out of the ground and pumped to the electric generation plants.

Apparently natural gas isn't easy to store, so basically there's not a lot of buffer, and everything needs to stay online or there won't be any natural gas to burn at power plants even if they were winterized, which some of them were.

I think some power plants may have even just shut down because natural gas was available but too expensive, and they weren't under a legal obligation to keep producing electricity.

Point being, to solve the problem, winterization of plants is just part of the problem.


The natural gas load was nearly double it's average load, because the entire state of Texas was below freezing for a week solid and almost nobody has gas heat in Texas. Everyone was pulling on inefficient electric heaters to warm their house. And the Windmills, which power about 22% of the Texas grid, went almost entirely offline.

“Almost nobody has gas heat in Texas” That’s incorrect. Approximately a third have gas heat per https://www.eia.gov/todayinenergy/detail.php?id=47116

Well, you must be a fun conversationalist. Nearly 2/3 of Texans use electricity for heat, and it no doubt becomes more electrically dependent in the southern latitudes of Texas (it being a very large and widespread state, climate varies quite a bit in various parts of the state).

By your own source that's 50% higher electricity dependence for heat than the U.S. average.

But no, I was not talking about the upper latitudes of Texas where freezing weather and snow is quite common and gas heaters are more prevalent.


You get plenty of people in Austin and more southern parts of Texas that also have gas heat. The older homes, typically — like from the 80s. Most newer homes are all electric.

Of course, if the fans are electric, then the gas heat doesn’t help.

In that case, you better hope you have a gas fireplace.


Natural gas plant outages were around 30,000mw, and there were issues with both coal and nuclear plants going offline also. Wind was only a few thousand MW below planned values.

Not close to being entirely a wind power issue. (Even though wind power could clearly have done better, as it did in Iowa, where it was significantly colder and the wind turbines stayed online.)


https://www.statesman.com/story/news/2021/02/14/historic-win...

Over half the turbines were completely frozen. Wind typically accounts for ~22% of Texas power supply. Meanwhile, there was a record high spike (nearly double normal loads) in demand for electricity. If not for natural gas, the blackouts would have been even worse, because at least they were able to be ramped up and provide some power via rolling blackouts.


> If not for natural gas, the blackouts would have been even worse, because at least they were able to be ramped up

Natural Gas, Coal, and Nuclear accounted for over 80% of the lost generation. Most of that was gas.

See page 16 here: https://www.ercot.com/files/docs/2021/04/28/ERCOT_Winter_Sto...

Same thing happened in both 2011 and 1989… NatGas was not there when the state needed it most.


The wind turbines need to be winterized, too.

They didn’t do as well as they should have, but they did a lot better than expected.


I recently heard a podcast with the CEO of MidAmerican (based in Iowa). They had materially colder temperatures than TX and operated 91% of their wind capacity. There's nothing inherent in wind that keeps it from operating in cold weather.

(MidAmerican did pay extra for winterization, so you can maybe make the argument that it wasn't worth it for TX to have done the same, given the risk profile.)


No, the windmills actually produced more power during that cold period than was projected that they would be capable of. They didn’t go offline. A standard Republican refrain is to blame alternative sources of power generation for the faults that actually exist primarily in the old systems that bought off said Republicans. It’s like the pot calling the kettle black, but the pot actually is black and the kettle isn’t.

No, it was the gas-fired power plants that didn’t have enough gas to burn, caused by the fact that many of the gas producers had not winterized their systems, and in fact some of the gas producers got their own power turned off in the rolling blackouts — they hadn’t been marked as mission critical.

Try reading the report.

Sadly, very little has changed since February. The producers and the gas power plants have paid a trivial fee that allows them to choose not to winterize (because that would be too expensive), and so far as I know they still haven’t even bothered to mark those systems as mission critical, so when the rolling blackouts happen again, they’ll get their power cut off again.


> And the Windmills, which power about 22% of the Texas grid, went almost entirely offline.

Which was also a problem of not winterizing.

The picture of the entire event is that the government of Texas flagrantly ignored what it was told was the problem from the previous blackout, and then naturally the exact same problem caused the next one.


This week-long freezing event had never happened before in the state of Texas. The entire state was blanketed and plunged to temperatures colder than Anchorage, Alaska. We had single digit temps in Austin. Austin, has the same latitude as Jacksonville, Florida. This isn't some "flagrant ignorance" this never happened in modern meteorological history in Texas.

https://www.texastribune.org/2014/01/06/ercot-rolling-blacko...

2014. After the event the government was advised that winterizing the powerplants needed to be done to prevent similar events in the future.

Instead, Texas did nothing for another 7 years, and then the same thing happened again.

This was wholly predictable: Texas running an independent grid doesn't have enough power import capacity to manage long-tail events, but also chose to manage them by ignoring them and damn the consequences - those consequences being Texan citizens froze to death in their homes (https://www.texastribune.org/2021/02/19/texas-power-outage-w...).


You might also want to read the analysis here around power sources during the outage: https://www.texastribune.org/2021/04/28/texas-power-outage-w...

For anyone interested, FERC basically just said this as well in their report on the outage.

https://www.ferc.gov/media/february-2021-cold-weather-outage...

Of course they said a bunch of things and made a bunch of recommendations the last time ERCOT fucked up, and they didn't listen to nearly any of them, so I would expect the same this time.


"The grid" tends to refer to more than just the transmission wires.

That seems unnecessarily sarcastic. I wasn't saying just the "wires". I was trying to shed more light on what the actual problem was.

Tbf, lots of people conflate transmission wires with "the grid". Even on a fairly tech-savvy forum like HN, I've read where people insinuate things like if all ICE vehicles were suddenly replaced with EV, all "the grid" would need is more wires running to charging stations.

Maybe people consider the grid distinct from all the supporting infrastructure.


“The grid” never refers to just the transmission wires.

I was surprised the power outage affected Samsung because in DFW the outage only affected residential areas. I didn't see a single commercial/industrial space without power including fully lit empty parking lots.

What finance guy will ever approve this? I doubt that the service agreement would ever be approved with that clause.

LOL, the current political heads for the state of Texas (and county of Taylor) will happily trade future liability of their constituents for investment now (and kickbacks later). Honestly, I think Samsung's management is more comfortable in this sort of "political" environment than many others.

If you watch local news reporting, it's hilarious that they (politicians) have no idea what the "factory" does. I'm personally highly doubtful that a fab which hasn't broken ground will be producing viable 5nm at the 2H of 2024 as claimed (or cyber trucks for that matter).


> the current political heads for the state of Texas (and county of Taylor) will happily trade future liability of their constituents for investment now (and kickbacks later)

Do you have some examples of this?


ERCOT, the massive statewide blackouts, and bailing out massive price gouging of the electric utilities that basically makes the reliability of the grid an issue in the first place? https://www.texastribune.org/2021/05/25/electricity-market-f...

Or, in general: https://www.texastribune.org/2017/08/09/are-business-ties-be...

Billion dollar "public/private" investments turn in to boondoggles except for those who retire as lobbyists or foreign firms who pick up the pieces of the bankruptcy.


You might want to do some research into the people in charge. You're kind of implying that those people would listen to bean counters over their pollsters. This is where the flaw in your logic is located. Also, you're using logic in the first place ;-)

Surely there's an insurance company that would take on that risk, and the government could pay for it. These government incentives are rife with excessive spending on a per-job/total economic impact basis. I see no reason why this detail would derail things.

Location close to ERCOT is not the reason.

Samsung for years has had a major facility outside of Austin.

Samsung Austin Semiconductor https://goo.gl/maps/nX9BFxHoVMCVKBLL7

And ERCOT is also based outside of Austin.

Austin is also the state capital, centrally located & home of one of the nations largest universities (UT), who has a strong engineering & CS program graduating thousands of students per year.

So it shouldn’t be a surprise that major companies have a significant presence in or around Austin.


I really doubt proximity to the ERCOT center has anything to do with it. The new facility isn't close enough at all to the ERCOT center to be on the same grid segment. And with a new facility like this that is in the middle of a field outside of town, the new fab will be built on its own grid segment anyway, completely separate from the rest of Taylor and the ERCOT center.

Taylor was chosen because that area (northeast of Austin, just south of Taylor) is one of the last areas near Austin that has abundant amounts of wide open fields ready to be developed. They also got a good deal on the water usage, and of course the tax breaks from Williamson County.


Do you have any thoughts on how water availability will be in the years/decades to come and how a fab would impact it and the communities depending on it?

My understanding is when you cross east past round rock towards Hutto/Taylor, the water aquifer being used changes to use water piped over from Carrizo-Wilcox aquifer.


Newer fabs like this recycle a huge portion of their water. I did some back of the napkin math a few weeks ago and the amount of water that the fab will be drawing (after accounting for the recycling) will be about the same as the water usage of an equivalently-sized suburban neighborhood. So while water usage is definitely a concern, I'm not sure that this fab is a bigger concern than your typical neighborhoods or farms.

tangentially related bit of trivium... when i was house hunting a looked at a house across from a fire station, and upon doing research about what it's like to live across from one, found that a.) they tend not to blast sirens upon exiting the station, in contrast to living near a hospital with an ED, and b.) the power grid segments that fire stations are on have priority and are less likely to lose power during blackouts.

I wonder if the siren thing varies from location to location, a good friend of mine lived right next to one and it seemed you couldn't go 15 minutes without being drown out by them heading out.

I wonder if the siren thing varies from location to location,

In my experience (former firefighter here), it absolutely does. It also varies by call type, time of day, etc. So if we got a call at 2:00 AM and there was no traffic on the road near the station, we might leave the station with no siren out of respect for the people trying to sleep in the homes very close to the station. This would be more true if the call was a lower priority call in the first place. OTOH though, if the call as "residential structure fire with occupants reported trapped" our focus would not be on sleeping neighbors and there's a better chance that we'd be getting on the Q pretty much right out the door.

Another factor for us was that our station was right next to an intersection that was very busy at times, and known for many traffic accidents. So any call where we had to turn left (towards the intersection) out of the parking lot, there was a better chance we'd be hitting the siren and air-horn pretty much right from the jump.

Anyway, yeah, net-net, this is going to vary based on lots of factors: department policy / culture, geography, time, traffic, etc, yadda yadda.


>getting on the Q pretty much right out the door

Q?



My office is adjacent to a fire station. The policy has changed over the years. They used to use the sirens all the time. Now they almost never use the sirens when they leave. In a way that makes it more disturbing when they do, you know there's a reason.

Other things which are bothersome about a firehouse…

• They put the communications radio on some sort of loud speaker sometimes, so you get to listen to all the calls in the city.

• They have some sort of turbine that runs a lot. Maybe a hose dryer? I haven't gone to ask, but when I used to do audio recordings there were days we couldn't record because of it.

• Their backup generator optimizes efficiency by minimizing the muffler. The tests are pretty loud.

• The building department made me block up the windows on their side because I was too close to the property line and a fire might spread to the fire station. I say if the fire department can't keep fire from coming out my window and setting fire to their building 60 feet away, that's at least partly their fault. :-)

• They are a multi-tier bureaucracy. When their tree slowly fell over on to my building it took many months to get them to admit it was their tree and maybe they should cut it down. Lots of finger pointing between the fire department and various city levels and a few "dead parrot" conversations with different functionaries.


If the fire station is located on a busy road or near a blind curve, turning on the siren when exiting might be a good policy. Out on a county road with good sight lines and low traffic? They can wait until they're going at a hazardous speed or intersection.

My main concern would be all of the communications equipment they typically have. Being that close to high powered transmitters can cause annoying interference.


Not just a good policy -- my wife is a former EMT & firefighter in MN, and at least there, apparently it's the law that fire trucks blast their horns at every intersection. Whether this is required when they leave the fire hall and turn onto the first road, I don't know.

A conversation may yield better too. If they know they are waking a day sleeper, for example.

I know I would attempt one. Try nice and some charm and see what comes of it. May end up popular with neighbors.


He's since moved, I'm not sure if he had spoken to them, but I think you're right that it would have made him a hero of the area. This was a pretty densely residential area of Waukesha, WI - the fire department near the airport.

Well, maybe a passerby benefits from this exchange.

In NJ they go off all the time to alert the volunteers not at the station, and it's used for all emergencies, not just fires that need trucks to roll out.

I live in NJ, directly across the street from a fire station. I also occasionally work nights to connect with team mates in other time zones.

You will always see the flashing lights but they have never blasted the sirens at night.


I lived in an area that lost power every time there was a stiff breeze. This was back in the 1990s, when we still used telephone modems. I blew up 3 of them. 2, even with surge protectors.

But power came back quickly, because I was on the same trunk as the fire station.


I used to live a mile or two away from a hospital and getting the lights back on before everyone else was nice.

> the power grid segments that fire stations are on have priority and are less likely to lose power during blackouts

So do large grocery stores, for obvious reasons.


It has nothing to do with ERCOT presence, despite the social karma of referring to the “grid disaster.”

There have been literally hundreds of other grid disasters in New York, Chicago, the entire state of California, and yet no one would be surprised if the plant showed up there. Why?


Isn't Samsung big enough to simply ask permissions to build their own nuclear reactor at that point ? Or wind turbine farms or whatever ?

A new nuclear plant cost billions and takes years to get permission and construction to complete. I assume they rather want to make chips, soon.

Well, for the sake of entertainment let's see: I suppose Samsung still want to be able to manufacture products in its Texan factories for more than 5 years (which according to my small European contry is roughly the time it took to build a nuclear power plant in the 70's) and they have billions (oh and Apple too) and according to Google's first results:

     Companies that are planning new nuclear units are currently indicating that the total costs (including escalation and financing costs) will be in the range of $5,500/kW to $8,100/kW or between $6 billion and $9 billion for each 1,100 MW plant.
So why not ? They'd control one more link in the supply chain.

"So why not ? They'd control one more link in the supply chain."

Because operating a nuclear reactor is a whole different thing, than operating a chip plant and you usually want to focus your energy, not spread it out. And when the goal is, to just have blackout save energy for your chip plant, then there are way cheaper, riskfree and faster solutions, like batteries, or gas generators.


Reactors and wind farms still need a grid.

Not if they are only interested in getting power to their factories ^^.

They are setting up fairly close to Tesla. Wonder if they are working on a deal for Tesla to provide battery backup. Interestingly enough Austin, TX appears to be rapidly approaching SF status as a tech hub and may soon surpass them. Its an interesting thing to watch and really gives insight into how politics and taxes affect business and growth.

> Wonder if they are working on a deal for Tesla to provide battery backup.

Running an entire leading edge semiconductor plant on battery backup is a crazy proposition in my view. Especially, for intentions of surviving another Texas winter situation without any disruption to operations. Approximately 24 hours of full-coverage battery backup would have been required to bridge the rolling blackouts.

Last I checked (about a decade ago), the SAS A2/S2 lines accounted for ~13% of the electrical load for the entire city of Austin. Just look at the amount of power one EUV light source requires today. These lines will have potentially dozens moving forward. Then consider that all of the photo area accounts for a tiny fraction of total power consumption in one of these plants.

If it were even remotely feasible for the Samsung semiconductor lines to operate on standby generator power, they would have installed these units already and no losses would have been incurred.

They might as well be smelting aluminum behind those walls. Any backup generator power or UPS devices are designed for life safety and stopping the line without causing 10+ figure losses. You cannot run one of these facilities on in-house power. Another commenter proposed a nuclear reactor installation. This is actually not a terrible idea once you understand the scale. Putting 10-15 semiconductor lines around a nuke plant makes perfect sense to me.


The Bay Area had something like 20x as much VC investment as Austin as of 2017 [0]. This article [1] indicates that the trend is probably in Austin's favor, as I'd expect, but it also shows that Austin isn't even in the top 10 nationally (unless they lump it in with San Antonio?) by deal count as of 2020.

[0] https://pitchbook.com/news/articles/the-bay-area-beyond-rank...

[1] https://www.cnbc.com/2021/01/14/silicon-valleys-share-of-ven...


> Interestingly enough Austin, TX appears to be rapidly approaching SF status as a tech hub and may soon surpass them.

Uh, maybe it's time to take a drive through Austin, but that was a pretty sleepy little college town 10 years ago. Most of my family lives in the central Texas region (San Antonio, San Carlos, College Station), and I presently live in Mountain View. I have a hard time believing SF-scale infrastructure was built that fast.

As an example, here's the UT wind tunnel: http://research.ae.utexas.edu/floimlab/Mach5.php

Here's the Moffett wind tunnel complex: https://www.nasa.gov/centers/ames/multimedia/images/2005/nfa...


+1 for comparing wind tunnel capacity. Also TX has no dirigible hangars.

Can't stop laughing

Companies that require wind tunnels are no longer “tech companies” and are therefore irrelevant for a discussion about “tech hubs”.

Yeah but San Antonio's even better, here's their car factory: https://s.hdnux.com/photos/46/41/53/10099417/69/1200x0.jpg

I think he may have meant in terms of software companies. Many of them do not require a wind tunnel.

I primarily meant software, computer manufacturing (chips, etc. ) style jobs but the fact that you went with wind tunnels is awesome :) You will always be cool in my books for this comment.

I don't think wind tunnels (of which these seem like fundamentally different types) are a great example for this argument.

I don't know about SV status but I think it's getting there. Other tech hubs still extract graduates out of the state at an alarming rate and last I checked compensation wasn't close to SV (even with normalization for cost of living). It'll take a good bit of time to undo SV but the ball is rolling.

care to share any sources? or did you just make this up?

Make what up, The Tesla link? I just asked the question, I have no info either way, notice I said "Wonder if".

> rapidly approaching sf status. That should be a quantifiable metric?

Austin has seen a 40+ % increase in tech jobs in a decade and has recently attracted major companies like Tesla and oracle to headquarter there while the majority of FAANG companies are building large campuses there.

This site does a pretty good job: https://sfciti.org/sf-tech-exodus/

Article from the SF Chronicle: https://www.sfchronicle.com/bayarea/article/Austin-COVID-tec...

Hopefully that addresses your question as to why I hold this opinion.


Yes, they made it up

According to wikipedia/some google-fu SV has just above double the high tech jobs as Austin


That would be way more high tech jobs in Austin per capita.

According to wikipedia/some google-fu Silicon Valley has a population of ~ 3 million people while Austin has a population of just under a million. Your numbers support my comment that you say I "Just made up"

I'm a big fan of Tesla and Mr. Musk - but does anyone know of any non-vaporware large scale industrial facilities that use Tesla battery walls as power backup? Do Tesla manufacturing plants do this?

Not sure about industrial, but they have grid scale battery storage with their Megapack product. Hornsdale in Australia has been operating for a few years. https://www.greentechmedia.com/articles/read/tesla-fulfills-...

https://www.bloomberg.com/news/features/2021-03-08/tesla-is-...

Not quite what you were asking about but it does deal with the TX grid.


What are some interesting startups in Austin area?


I doubt it. The land is just cheap and widely available over there. Also, they know workers will be attracted to being a short commute from Round Rock. The storm that caused the grid failure is a very rare occurrence ("100-year storm").

Which means that it will probably happen this year, too. And every year after this, thanks to the global catastrophic weather changes.

Doesn't it cost less than that amount of money to have on-site backup power?

If you are that big of a deal, wouldn't you have your own power plant with reserve power? My university had 2 power plants. Surely it isn't that hard

Why not use generators like most data centers in case of power failure? Is energy usage that much higher?

As more and more companies start to invest in Texas there is more demand in tigetening the grid to avoid future catastrophes.

I worked at Samsung Austin Semiconductor for ten years as a software engineer. AMA.

I very much love the people I worked with at SAS. It was one of the best experiences of my life and I'm very happy to see them expand like this.

Wayne at Louie Mueller's BBQ had better add a few more pits, they're about to get a lot busier!


How do you get into chipmaking, really? Is it a Physics degree, or is it a Computer Science degree with just lots of on-the-job training?

It depends on what you want to do.

If you want to work on one of the processes in the fab (metrology, chemical vapor deposition, photolithography, etc) then an engineering degree (especially chemical engineering) may be your best bet.

If you want to design the chip architecture, then computer engineering is the degree path you want. Samsung has their own chip design subsidiary in Austin too, over on Highway 360 in west Austin, if I remember correctly, called SARC.

Austin Community College has associate degree programs designed for technicians.

I was a software engineer responsible for many of the internally-facing enterprise systems, as well as the intranet. I worked with all of these engineering departments on many different projects, so I had exposure to and visibility into many functional areas and processes of a truly global, high-tech manufacturing concern. In one of the random ways the universe has rolled the dice in my favor, one day a recruiter called me and asked, "Would you like to work for Samsung?"


I know a few people who went to university near a large Intel campus that was always hiring.

These people told me that if you got hired into Intel, it was almost always as a design verification & validation engineer, and that getting an actual design (architecture) job was near impossible (especially as a recent graduate, even with an MS).

Is it the same for Samsung?


In the Austin area for Samsung, there is an order of magnitude more staff dedicated to verification and validation as part of the manufacturing process than chip design. My perception was that the chip design folks (SARC) are like the golden children of the ecosystem. Their salaries reflect that it's obviously a more competitive, demanding environment. I don't know how hard it would be for a fab engineer to transfer positions from SAS to SARC, but I do remember that it was not a common occurrence.

how could a new grad be qualified to design a new intel chip?

seems like those are necessary stepping stones to learn all the real world issues that would be involved in actually designing a chip.


> how could a new grad be qualified to design a new intel chip?

How could a new grad be qualified to design a new Google tool?

By being an apprentice (intern) at Google who is initially only responsible for a small bit of code (not "the tool" nor "the chip" - just a piece), and whose contributions are thoroughly reviewed and tested before pushing to production.

This comment isn't very constructive.


If they're talented and knowledgeable enough to contribute. Top tier students can do it. They don't need to start in validation.

There are two validation engineers for every design engineer in the semiconductor industry. There are even fewer design architects. But they all need a computer engineering background. Design verification is just what’s in demand.

I should add, if this is a path you're seriously considering, I will get you in touch with people at SAS, either in HR or in the engineering departments who can offer guidance. Samsung Austin Semiconductor has community engagement programs as well as career days for Austin-area high school students.

Samsung also recruits directly from Texas universities. I remember many Longhorn and Aggie engineers being hired straight from school every year. I don't know what your school situation is like, but I can also get you information about how that pipeline works.


This is just my experience working at another Big 5 manufacturer, but the easiest route was entry from another engineering role (if you didn't know anyone on the team there already). They never had any undergraduate roles available as the pipeline generally came through at MSc or PhD level.

I started in application engineering with an EEE degree, built a name for myself there over a couple of years, and made my intentions known once I built a good repertoire with my manager and mentors that I wanted to transition into chip design. They helped me network with the right people, and the chip design team took me on. Some of the skills I learnt at university, most of it I learnt on the job.

Networking > Qualifications to be honest. On the team we had Physicists, Electronics Engies, Comp Scientists, ML Engies, Mathematicians, Chemists and so on. Provided you can get a foot in the door and convince them you've got some skills that would be of use, then you're good.


Undergrad in electrical engineering at a school that teaches coursework in VLSI and computer architectures. There are a few that have professors and course studies at the undergrad and graduate level in semiconductors, integration with partnerships with industry. Off the top of my head, UIUC, UC Irvine, UC Berkeley, NC State, UF, UT Austin, Cal Tech. I'm sure there are many more.

Graduate degrees in physics and EE focused on semiconductor physics and manufacturing are also useful. I've known a few post grad/doctoral folks from nuclear engineering and optics that have careers in the field.


I studied what is formally known as Computer Engineering at Oregon State University.

The courses we took had labs starting from soldering physical components on boards (actually wire wrapping) to hardware design tools.

After graduating, I went to work for Mentor Graphics, (now a division of Siemens). They create chip-level design tools of all kinds, the big moneymaker back then was Calibre Parasitic Extraction.

They also sell PADS, a PCB design tool that was popular. I did lead gen for both these products and steered away from design and into software.

However, you could go to work for any of the electronic design automation (EDA) companies and ultimately learn the design tools at great detail which might be a way to jump into a position on a design team.

At the very least you would be in a position to build a professional network with chip designers.


At least on the hardware side of things, it's an Electrical or Computer Engineering degree, usually with at least a Masters

>I very much love the people I worked with at SAS.

What color is the boat house at Hereford?


update hereford.buildings set color="blue" where building_name="boathouse" or 1=1;

I reckon it's blue.


you forgot to commit the changes. surely, you don't operate with auto-commit on in prod do you?

Sometimes we programmers just need a little danger in our lives.

  function main(code) {
    try {
      eval(code);
    } catch(e) {}
  }

main('Bobby drop tables')

Good to see a fellow (former) Samsung employee. What company did you move to? As I've also primarily worked on in-house software only, I'm finding myself less and less competitive in the era of venture capitals and IT services. Did you find a solution?

> What company did you move to?

After working for Dell and some small start-ups for a while, I now help the federal government in various ways. I started working remotely full time in 2017 and live in the Hill Country. This is me.

https://www.linkedin.com/in/jason-phillips-2505a0187/

> As I've also primarily worked on in-house software only, I'm finding myself less and less competitive... Did you find a solution?

Yes, I had to solve this myself. Study, study, do side projects and then study some more. You have to become an expert in something on your own time. Whether that's full-stack software engineering, DevOps/SRE or whatever, spend six to twelve months studying and creating side projects. Create an AWS account for the free tier and get after it. When you apply, make sure to demo and speak to your side projects. In 2015-2016, I learned React, Angular, Nodejs, studied Kyle Simpson's "You Don't Know JS" series, Docker, AWS, and Python. That, combined with my enterprise full-stack development experience, was good enough back then.

I think it's easy for the grass to look greener somewhere else from where you're sitting, but I can tell you Samsung took really good care of me and was actually better run than most companies for whom I worked afterwards. Feel free to reach out if you have any questions.


It is great to hear about your experiences on what to do to overcome initial career choices. I agree that I might be seeing the grass in the other places greener. I'll have to be more careful when considering moving to other companies. Thanks for your kind and thorough answer!

Louie Mueller was really great. Never been to Franklin's but I can't imagine it's too much better.

Indeed, they will likely be busier.


Ohh emm gee, when I worked at SAS I dragged my coworkers to every BBQ joint in Austin. I think our favorite became Valentina's Tex Mex BBQ when it was still on West 6th.

Franklin's is a religious experience, imo. I went there several times with some SAS co-workers back in ~2013 when Aaron was still the cutter. Great times.


‘Impossible':TSMC founder Morris Chang on US dreams for onshoring chip supply chain

https://www.taiwannews.com.tw/en/news/4327937

Morris Chang spent 25 years (1958–1983) at Texas Instruments.He knows US very well.


Having worked in Fab, the supply chain is already in the US+EU+Japan. About the only change in last 5 years or so is Intel's ability to compete in lithography process and get sufficient yield to stay competitive in pricing. Lithography is one of the 400 or so steps in Fab processing, but arguably the most critical and a bottle neck. TSMC does not have a lead in every single of these 400 steps, in fact quite the opposite.

Anytime someone says "impossible", I would take that opinion with a grain of salt.


According to US Semiconductor Association's recent State of the Industry Report, Taiwan's share in the manufacturing equipment part of the supply chain is almost nil and entirely depends on suppliers in the US, Japan and EU.

  Manufacturing Equipment (12%):
  US: 40%
  Japan: 32%
  EU: 18%
  S Korea: 4%
  Taiwan: <1%

  Wafer Fabrication (19%):
  Taiwan: 20%
  S Korea: 19%
  Japan: 17%
  China: 16%
  US: 12%
  EU: 9%
I guess Apple's decision to leave Samsung's Austin fab back in early 2010's to TSMC in Taiwan had an outsized role in TSMC's lead in tech and rise in revenue -- Apple accounts for 25% of TSMC sales.

[1] https://www.semiconductors.org/wp-content/uploads/2021/09/20...:


There is also Software/IP which is basically US-based. For example, MES 3000 runs the fab which I believe is developed by AMAT.

> Anytime someone says "impossible", I would take that opinion with a grain of salt.

Or to quote Arthur C Clarke:

> When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.


Or, morbidly from Planck, "Science progresses one funeral at a time."

> Lithography is one of the 400 or so steps in Fab processing, but arguably the most critical and a bottle neck.

In their ongoing series on supply chains, the Odd Lots podcast had episodes on both TSMC and ASML:

* https://podcasts.apple.com/us/podcast/the-story-of-how-tsmc-...

* https://podcasts.apple.com/us/podcast/asml-the-obscure-power...


You have to take his words with a grain of salt. Without going into politics, it’s impossible to talk about TSMC without geopolitical context.

It’s essential to TSMC survival for their production to physically stay focused in Asia.


Morris Chang has been exceptionally clear on this issues when he was still CEO of TSMC, even before the whole politics and TSMC becoming leading edge.

It is impossible with respect to current cost structure, without substantial government money, while getting the same margin. Of course it is entirely possible to throw money at the problem and succeed. Which is what Intel is trying to do right now with the help of US government.


> Morris Chang has been exceptionally clear

He can be "exceptionally clear" and still biased in favor of TMSC.

> It is impossible with respect to current cost structure

Well, we'll have to pay a few cents (or dollars) more per chip then. It is infeasible for us to continue to depend on TMSC in the long term - it will be taken over by our largest geopolitical adversary by 2030 at the latest. We do not have a choice but to either on-shore, or shift production to countries with far less geopolitical risk.


On the other hand, from the perspective of TSMC, if production does not get transferred to USA mainland, then it secures the future of TSMC because in that case USA will ensure that they do not get taken over by a geopolitical adversary in the coming decade. Geopolitical risk is not something external that exists in isolation; the location of semiconductor manufacturing is a big factor that shapes the geopolitical risk by shaping the interests of key players. If USA can afford to walk away, Taiwan is at large geopolitical risk; but as long as their interests are tied together, then the risk to Taiwan is greatly limited.

> in that case USA will ensure that they do not get taken over by a geopolitical adversary in the coming decade

How will the USA ensure that?


If the push comes to shove, a US carrier group parked in Taiwan's territorial waters makes any sea invasion impossible if it's willing to shoot at Chinese ships; you can't ship and land a million soldiers across a hostile sea, and China can't (at the moment) win the sea battle.

It's debatable to what extent USA is willing to fight for that, but if it chooses to do so, then now and at least until 2030 USA has enough military might to enforce a stalemate/status quo across the Taiwan strait, no matter how much it gets escalated. China can force Kinmen and Matsu islands, but invading Taiwan proper requires either USA not joining the war or a stronger China.

So IMHO the fate of Taiwan in this century depends on how much USA will be willing to intervene to protect it; and the physical location of semiconductor factories is a big factor influencing that willingness.


If a carrier is willing to shoot at Chinese ships and close enough to reliably do so, then the Chinese are willing to shoot at it and are close enough to reliably do so.

Chinese antiship weapons are significantly better than the US. Without immediately inducing Kessler syndrome and still being very lucky, I don't see how a US carrier is going to survive two dozen DF-21 missiles simultaneously, a couple hundred supersonic cruise missiles, and whatever submarines or UUVs the Chinese have got lurking there.

The US fleets most advanced weapon, SM-6, still can't reliably intercept MRBMs or SRBMs(https://www.defensedaily.com/mda-conducts-sm-6-missile-raid-..., https://www.reuters.com/business/aerospace-defense/us-fails-...) and has never intercepted a single IRBM. I struggle to see how it can possibly defend against dozens of IRBMs with advanced countermeasures that aren't on a predefined trajectory if it can't even hit a single one with no countermeasures. And by the way, they're building over 100 missiles a year. They could realistically have 1000 antiship IRBMs in a few year - more antiship IRBMs alone than the combined amount of interceptors of all carriers the US could hope to deploy within 8000km of Taiwan.

Right now the best the US can do is try to jam the missiles and deploy smoke, and hope they miss, which they probably won't.

A US carrier group in Taiwan's territorial waters is 100% getting absolutely wiped. The US will never risk getting their carriers that close to Chinese weapons, they'll stay thousands of kilometers away and hope for the best.

The US's best hope is to keep its assets far away from the theater and cooperate with Taiwan to try to whittle away the Chinese fleet and prevent them from setting up a beachhead, and hope the Chinese air defences, low-frequency radars, and signals intelligence platforms can't stop the counterattack.

The Chinese also won't need a million soldiers. They just need to get a few dozen thousand on the shores with air superiority and disable the Taiwanese military then take as much time as they want pacifying the island.


In a full scale conflict, carrier group would be sunk by subs in the first 15 minutes. Those carrier groups are a relic of the past that only works against unsophisticated adversaries who do not have the largest submarine fleet in the world [1] or nukes. US military planners know this so any such things will be withdrawn shortly before shit is about to go down, or (more likely) never brought in in the first place. And it doubly does not work against adversaries without whom we can't even build anything because we outsourced all of our production there. In 2 years we couldn't even scale the production of N95 face masks on US soil, let alone anything more complicated. The colossus has legs of clay and US government (and its owner - the US business establishment) is to blame.

Prediction: Taiwan will be retaken by China by 2030 and the US will do bupkis about it other than saber rattling. It can't even do sanctions, since that'd be sort of like US imposing sanctions on its own manufacturing base. Xi knows this. He will use our weakness to his advantage. It would seem that business establishments are already acknowledging this, hence this massive investment in the US. On the other hand, this is also forcing China's hand - they can't afford to wait until it becomes feasible for the US to impose effective sanctions. Their colossus too has legs of clay. Except the legs are not as weak.

[1] https://www.globalfirepower.com/navy-submarines.php


To those downvoting, if your attention span is long enough for a long read, here's an explanation of why the United States will 100% guaranteed lose if it chooses to engage: https://scottlocklin.wordpress.com/2021/11/03/america-agains...

I read that. Doesn't at all explain why the US would be guaranteed to lose in a naval conflict in the Taiwan strait.

The US has no hope to even possibly defend surface naval assets in the Taiwan strait or its periphery. A carrier strike group literally doesn't have as many interceptors as the Chinese could fire missiles at them from that distance.

The US will have to leave the Taiwan strait waters to China and attempt to strike as many Chinese boats there as they can from a distance in the air.


Because the US has been ruled by incompetent and corrupt imbeciles since the 70s. And there's now a second generation of incompetent imbeciles in the government who were students back then. If you think they can avert, let alone win this, you're out of your mind. Our best play here is to not intervene on another country's behalf seeing how we couldn't win against a bunch of semi-literate cave dwellers with Kalashnikovs after 20 years and 2 trillion dollars. War against today's China would be unwinnable in _any_ shape or form whatsoever under any circumstances even if we had competent leadership, which is something we do not have, and haven't seen in 50 years.

Watch the YT video linked in the article. China is in the position of strength now. They're on an upward trajectory. We're on a steep downward one. And that won't change until we at least acknowledge the realities of our present predicament, and begin the serious, careful work to rectify it. I see no signs of that even being possible the way things are right now, let alone likely.


I suppose "ensure" is too definite of a claim. But it's clear that the US can apply all types of pressure to try to hinder a takeover of TSMC/Taiwan. Methods range from open negotiations, to trade/economics sanctions, etc.

>He can be "exceptionally clear" and still biased in favor of TMSC.

Probably. But It wasn't that TSMC doesn't want to move, he made it clear the maths doesn't work out. Giga Fabs operate at scale and efficiency is precisely why you could make a $30 M1 Die.

>Well, we'll have to pay a few cents (or dollars) more per chip then.

That is easier for mature node. But most of the time we are not talking about mature node but leading edge.

If the maths were a few cents or even a few dollar things would have been done already. In US, for TSMC having the same margin would probably put M1 close to $50, while having higher initial R&D cost, meaning more volume to amortise the cost, or higher final retail price depending on sector.

So Morris's question to TSMC's client, are you willing to pay 50% to double for US made silicon. As far as all major US players, that is Qualcomm, Nvidia, AMD and Apple. The answer has been a simple no. They want it cheaper! ( Looking at Nvidia's constantly moaning )

Of course there is another path, forcing TSMC to operate in US while lowering their Net Profit margin. Which is likely what is happening here. The only good thing is that US is finally understanding the risk of its supply chain. For some people like me who has been crying about it for nearly a decade.


OK, then we'll pay $20 more for a $3K laptop. Hardly a tragedy. Same with labor by the way. If Apple made phones here and charged $20-30 more per unit, their user base wouldn't even blink.

I guess I have to be explicit.

For a $3K Laptop you are looking at M1 Pro or M1 Max, those would cost at least $50 to $100 more BOM cost. Excluding other factor. That translate to roughly $125 to $250 retail price increase at the same margin.

For a labour of $20 to $30 increase per iPhone unit, would equate retail price increase of $50 to $75. Excluding other factors. Not to mention I seriously doubt labour would be an increase of only $20-30 per unit. The different in cost / productivity is likely higher than 3x.


There's not a whole hour of labor in each iPhone. Not even close. Humans just put together robot-assembled boards, screens, and so on. If this is re-shored, it wouldn't be that difficult for Apple to reduce manual labor even further than that, possibly to near zero. I've tried to repair an iPhone PCB once where an SMD resistor fell off. I couldn't do it, even though I have a hot air workstation and a bench microscope. Humans can't work on SMD components you can't even see without magnification.

And I'd like to understand how it is that an M1 Pro die, which is manufactured _solely_ by robots, and can't even use human labor would cost $100 per unit more here than anywhere else in the world. Heck the chip itself likely costs less than that to manufacture (do remember that we're excluding the design cost, which is the same).

This is quite literally the case of big businesses selling their (for some tenuous definition of "their" - they don't much care where they are situated as long as they have access to US market) country out for a buck.


> an M1 Pro die, which is manufactured _solely_ by robots, and can't even use human labor

I was not aware that there were no employees in modern chip fabs. They must require some sort of maintenance cycle, how often do humans come to modern chip fabs?


"Maintenance" won't add $100 per die like GP is suggesting. There are hundreds of dies per wafer, and thousands of wafers within maintenance interval. We're talking a very small incremental cost. And it too could be reduced to account for more expensive labor.

I'd like to pose a better question: why do you voluntarily carry water for Apple? If you are in the US (or in one of its allied countries), on-shoring advanced semiconductors, and other types of advanced manufacturing is 100% beneficial to you no matter from which perspective you consider it.


It's essential to Taiwan's survival for production to physically stay focused in Taiwan. I don't know if the US would be as willing to defend Taiwan if our own recovery wasn't affected by the chip shortage.

Yep. to use a quote I've heard about another resource: "No one would care as much about stability in the Middle East if they produced toothpaste instead of Oil".

Unfortunately at the same time for Taiwan, once the West's reliance on that area of the world for production becomes less of an issue, China's timeline gets accelerated much faster to the "No but really you're part of the PRC" point. Similar to how the PRC ratcheted up control over Hong Kong in 2020 when the rest of the world was basically "Sorry HK we got COVID issues k'bye"


> Similar to how the PRC ratcheted up control over Hong Kong in 2020 when the rest of the world was basically "Sorry HK we got COVID issues k'bye"

Little more complicated than that - with HK there wasn't much the west could really do, or any obligation to do anything at all. There were no treaty obligations, as far as I'm aware of.

Obligations to Taiwan have been covered since 1979, and there's the whole matter of the PRC needing to do am amphibious invasion against a prepared foe.

I'd argue they are not similar at all, really. Similar only in PRC's ambitions.


Once the chip shortage happened, partially due to geopolitical reasons, that changed everything. The West/US will never fully rely on a single point of failure again no matter how hard the HBS MBAs and Chicago thinking push it to trim and be "efficient". Some industries are too important for other industries and leverage on that over those areas is too risky and costly now.

I'd pay double right now for GPUs directly from the source, not from some sketchy third party.

Right now our EV/auto, space and even AR/XR industries as well as gaming and everything that requires chips, we are at the mercy of an external market that has a slant against the West. It will take some years to get out but we'll never not expect that in the future again. If costs go up costs go up, but availability should never be allowed to be used as leverage again, that is too risky and too costly long term.

Availability that is reliable is always more important than efficiency or cost, because right now lack of availability is costing lots of extra time that has the potential to lose entire industries, that is not acceptable in any way.

Very little margin and too much optimization/efficiency is bad for resilience. Couple that with private equity backed near entire market leverage monopolies/duopolies/oligopolies that control necessary supply and you have trouble.

HBS is even realizing too much optimization/efficiency is a bad thing. The slack/margin is squeezed out and with that, an ability to change vectors quickly. It is the large company/startup agility difference with the added weight of physical/expensive manufacturing.

The High Price of Efficiency, Our Obsession with Efficiency Is Destroying Our Resilience [1]

> Superefficient businesses create the potential for social disorder.

> A superefficient dominant model elevates the risk of catastrophic failure.

> *If a system is highly efficient, odds are that efficient players will game it.*

Hopefully that same mistake is not made in the future. It will take time to build up diversification of market leverage in terms of chips for availability. Hopefully we have learned our lesson about too much concentration, with that comes leverage and sometimes a "gaming" of the market.

This chip shortage, and all the supply chain problems during the pandemic as well, will hopefully introduce more wisdom and knowledge into business institutions that just because things are ok while being overly super efficient, that is almost a bigger risk than higher prices/costs. Competition is a leverage reducer. Margin is a softer ride even if the profit margins aren't as big.

[1] https://hbr.org/2019/01/the-high-price-of-efficiency


The Taiwanese gov't was deeply involved in TSMC from its inception in the 80's to what it is today. I'm guessing it's in their best economic, geopolitical interest to keep things in Taiwan as much as possible.

Person who runs billion dollar business says competitors will fail, encourages people to continue to rely on him. News at 11.

key phrase:

> a full chip supply chain in the country

(emphasis mine)

TSMC is, after all, ALSO building a fab in the country, so it's not like TSMC doesn't believe in onshoring chips to the US. The fabs are {high capital expense/high recurring expense/high skilled labor} operations so the US is very competitive due to the labor value-add. Not every part of the full supply chain is like that.

To serve (aka reduce risk to) US supply chain interests in the short term, you also don't need access to the full stack domestically. You could potentially send a lot of activites that need low-skilled-labor to, say, Mexico, which is a much lower supply chain risk that Asia (aka thailand, vietnam, china). National security onshoring, which possibly has less tolerance for even outsourcing to Mexico, is likely a different story and more aligns with Morris' statement.


I also think people get focused on having a state-of-the-art production. Lots of value in having a domestic facility that can make last N-tier generation chips (10, 14, 22, 32, etc) are all capable of providing significant value.

This was mentioned in the last meeting held by Biden with the North America group. The Mexican president said that the imports from China should be reduced as much as possible in order to stop China becoming more powerful that it already is. Mexico at least in the mid/near future is a safe bet in terms of supply chain risk.

Maquiladoras (near the border) are even more resistant to supply chain risk

Well that sounds about right from the guy who moved it all to Taiwan. I hope the scene has changed a little in the last 40 years. The US is eager to have these businesses, and I can imagine a lot of incentives and regulatory reductions.

It has changed. That's what makes it impossible.

>dreams for onshoring chip supply chain

Depends on what the dream is. Bulk onshoring? Probably not while the geopolitical situation stays on the status quo.

But I don't think that's the attempt we're seeing here anyway. I think what we're seeing is the foundations built for the capability to bootstrap more robust manufacturing in the West if/when China decides that Taiwan really should stop claiming independence, and perhaps at the same time exerts its local power over that entire region of the world and thus South Korea as well.


Biggest change since then are relative energy costs between Texas and Taiwan. (To say nothing of geopolitical stability.)

quote is from oct 2021

This is a great first step. The more fabs that are built here, the more supply chains that feed them will follow. Supply chains tend to follow the manufacturers where feasible. This will eventually improve things even more for the other high tech companies in the state as they do move supply chains closer to production. Texas has been making some pretty substantial long term investments in the last few years. If you don't plant trees today, you won't have shade to sit beneath 20 years from now. The best time to start is often yesterday. These things take time and a great deal of capital and I'm glad to see some places are moving forward.

So everyone is on the "Let's build a Fab in the USA" train, will that eventually bring down the chip production costs? (not an expert in the field, so asking naively)

This is more about national security and business continuity than cost. Costs matter most when things are running smoothly. However, COVID has taught us a hard lesson. When SHTF, you better have a back up plan. Making money on IP alone while someone else handles all of the manufacturing and integration work looks great on a spreadsheet. Then some state actor cuts off the supply chain (or a virus) and your genius MBAs start to not look so smart after all.

With any luck, COVID (and tariffs) has transformed business for the future.


> With any luck, COVID (and tarrifs) has transformed business for the future.

Until we're far enough removed from the situation, everything is comfortable again, and the MBAs start saying "you know, if we just offshored fabrication..."


The pendulum swings back and forth similar to how politics have shifting tides that ebb and flow.

How is relying on a foreign company helping the situation?

Samsung is probably more accurately described as a multinational company in this context. The location of a company's headquarters is not a primary determining factor for all different types of geopolitical risk. Like other multinationals, they have some ability to pick the jurisdictions they operate, and each of those locations have their own geopolitical contexts.

However, in this case, South Korea is also a close US ally. There is little risk to the US that the headquarters of Samsung is in South Korea.


>However, in this case, South Korea is also a close US ally

So is Taiwan.


But TSMC isn't necessarily a multinational company, which can also be said about other notable Taiwanese companies like Foxconn, whose business model depends on exploiting young, cheap, unskilled laborers from rural China and who had been far less successful in operating oversea operations outside China, such as in Brazil or Wisconsin, US.

As insinuated in the pessimitic tone of in the interview, TSMC also appears to be very reluctant to expand beyond Taiwan and China, and is certainly less worried about geopolitical threat from China. I think it's important to note that not everyone in Taiwan is dead set against the CCP. In a pre-pandemic Pew Research Center poll[1], at least 1/3 of Taiwanese still longed for closer economic and political ties with the mainland China. I don't have any recent data, but I doubt it changed much since.

[1] https://www.pewresearch.org/global/2020/05/12/in-taiwan-view...


Correct. Which is the reason why having TSMC in Arizona would also help. https://tsmccareers.com/tsmc-arizona/

For these risks physical location matters more than the organizational structure. If something (perhaps a conflict, perhaps something else) disrupts manufacturing or shipping around East Asia, supply from the Samsung fabs in Korea and TSMC fabs in Taiwan would be disrupted, but any manufacturing capacity that's physically located in the West would be still available.

Well in principle Texas has lower (almost half) industrial energy costs than Taiwan, which could help reduce wafer production costs especially at the lower resolution nodes. As Moore’s Law slows, things like energy and raw material costs may matter more. And having wafer plants close to fab plants can help.

Rough order of magnitude, but for a lower cost wafer (think like 90nm or above, not 7nm), electricity and energy could be about 10% of the overall cost. As a lot of the capital cost of the automation and industrial equipment is the same wherever in the world you are, that might be an important factor.

Taiwan is an island that imports almost all of its energy. Texas is an energy rich state with lots of wind, solar, and especially natural gas that is exported to the world.

We talk about higher natural gas prices in the US with Texas seeing slightly above $4/MMBTU, but LNG prices in Asia are above $35/MMBTU at the moment, nearly an order of magnitude higher. The government of Taiwan is basically eating half the entire cost of LNG right now to keep costs low for businesses and consumers, but that’s not a sustainable strategy. (At the moment, industrial electricity is about 13¢/kWh in Taiwan even with the government subsidizing most of the LNG cost and 7-8¢/kWh in Texas.)


Why are people concerned about energy and material costs? Semiconductors are probably the highest margin manufactured goods on the planet, with materials and energy being the lowest cost input to the process. The only good that can compete is bottled water or movie theater popcorn.

This is 100% politically motivated.


It takes about 3 kWh per square centimeter to make a high-quality silicon wafer. 90 nm node silicon wafer might cost about three dollars per square centimeter. So it isn’t a rounding error for actual wafer production. The unsubsidized electricity cost in Taiwan is on the order of 30¢/kWh for natural gas vs about 7-8¢/kWh in Texas. So the difference isn’t zero.

Purifying and growing high quality wafers is very, very energy intensive. This becomes more important as Moore’s Law slows and more of chip costs become just the wafer costs.


I was wondering why Texas keep coming up when new US fabs a being talked about. From an outsider it just looks like the low energy cost comes from failing to properly manage and upgrade the Texas energy grid.

The Texas grid has been diversified and transmission upgraded. The problem is that it hasn't been winterized. While this is costly, I doubt this is the main reason why prices are low. Texas has a large, cheap supply of natural gas and good areas for both wind and solar.

The energy prices are low because Texas has tons of oil and gas and wind. The energy grid is actually more efficiently managed than California, which has had similar problems with electricity supply.

But the main reason Texas comes up is it has a very long history in semiconductors with a big workforce and lots of semiconductor companies.


Efficiency in electrical grids can often be counterproductive to reliability.

It is economically efficient to not winterize your generation capacity or build longhaul electrical transport infrastructure to move electricity regionally, but without these rarely used elements you could end up with severe load shedding or a black start event.


Clearly. Texas needs winterization.

But of course, you can be both economically inefficient AND unreliable. California is one such example.


(Also not an expert.)

In general I wouldn't expect US manufacturing to do a lot to lower production costs. More capacity in general obviously helps, and I could definitely see the US doing a lot to subsidize domestic production, if that counts as bringing down the costs.

Maybe higher wages will lead to more automation.


> Maybe higher wages will lead to more automation.

Samsung Austin Semiconductor has been fully automated since 2008-ish. One number I was quoted was that you only need 19 technicians on the floor to run the fab, compared to 500+ pre-automation.


Expert here. Worked in a Fab (Automated Material Handling Systems or AMHS group).

You might be suprised how automated it already is [1]. Notice the OHTs (Overhead Hoist Transport)[2] running on a system of highways and superhighways on the ceiling. Each of these OHTs carry a box inside it called a FOUPs (Front Opening Unified Pods)[3] carrying upto 25 wafers. That can easily be worth $1M+ per box. They have standardized JEDEC interfaces so any process tool can accept a FOUP and get wafers out of it. The environment inside these FOUPs is ridiculously clean and each one costs $15k for what most people would think is a plastic box with a door in the front. FOUPs are used internally. If wafers are shipped, they're shipped in a similar box called FOSB or Front Opening Shipping Boxes [4]. These boxes are the reason automation happens with standardized "APIs". We don't need to know who made the process tool to hand them a box of wafers.

Another automation feature are the AGVs or Automated Guided Vehicles [5] that carry the same FOUPs and FOSBs [5]. They also can be equipped with a robot arm that takes a wafer out of the box and hands it over to the tool. The AGVs run on specialized roads inside the Fab and remains docked on the tool during processing.

Usually, there are hardly any humans around the tool. Material (wafer) is fed by AMHS systems, picked up by the same. Fab is probably one of the most automated manufacturing places out there. Way more than Automotive manfacturing.

If you're looking to do software engineering, I would highly encourage people to go work in semiconductor industry. It is just a fucking cool place and lot of room for improvement still.

[1] https://www.youtube.com/watch?v=MRlcZqqyBM8

[2] https://www.muratec.net/cfa/products/

[3] https://www.entegris.com/shop/en/USD/Products/Wafer-Handling...

[4] https://en.wikipedia.org/wiki/FOUP#/media/File:Front_opening...

[5] https://www.muratec-usa.com/machinery/clean-room/transport/A...


For someone interested in robotics and automation, what types of groups/jobs would one look for specifically in this industry? Also geographically where are these types of jobs located?

If you want to work on robotics, you should look into working for one of the major semiconductor equipment manufacturers, not a Fab. To name a few: KLA Tencor (US), Applied Materials (US), ASM, ASML (Netherlands), Advantech (US), TEL(Japan), Besi (Netherlands), etc.

thanks for this highly interesting comment about the fab floor.

So chip production does have a labor cost, but the equipment used is quite pricey, and so the relative cost of labor is less. Also, given how pricey the equipment and how demanding the cleanroom protocol (which, if not followed, reduces yield of saleable products), there are a lot of incentives to make sure you have long-term employees who are well trained and motivated to keep their job.

Also, Austin (and other parts of Texas) already have a fair number of chip fabs, and have for decades. I used to work in some of them. Samsung even has existing fab capacity in Texas. So it's not such a new thing.


No, it increases the security and continuity of our civilization when China invades Taiwan.

Well, there's a huge benefit of COVID.

The related chip shortage finally woke up the US gov on supply chains.

My question is why did it take this long? I have been worried about this for nearly a decade.

My next question is how many mask factories have we built in the US since we realized that was important?


>> My next question is how many mask factories have we built in the US since we realized that was important?

There's a guy with a small company that makes N95 masks in the US. He got pissed when they came looking for a production increase. He would be happy to increase production, but nobody was willing to buy from him long-term. They'll just go back to importing cheaper product (probably from China) after the crisis is over. He basically told them to go pound sand.

I just Googled this and it seems a bunch of companies tried to increase production: https://www.nytimes.com/2021/02/10/health/covid-masks-china-...

The original guy basically said nobody wants to pay for anything, and I think he's right.


> (probably from China)

Not only China, they most likely come from Wuhan where the pandemic started.


We shouldn't stop there.

America needs to make steel again. Plastics. Chemicals.

We need to make electronics, fertilizers, tools.

Everything we need should be onshored. Especially if we expect a cold war with our biggest producer.


The "imperialist" model can still work fine if we want to stick with it.

We can keep production in poorer countries with poor worker and environmental protections so long as those countries will clearly side with us and aren't all concentrated in China's sphere of influence. No one's doing a naval blockade of the entire US coastline. And even if they did, we also have some long land borders with allies.


The problem is that US needs to find responsible governments other than the ones close to China. There are not many in the world. They are poor for some reasons.

BTW when I say responsible governments I don't mean democratic, but mean they care about people's education, can keep long term policies and build good infrastructures.


> My question is why did it take this long?

In general, I'd say we've been able to deal with issues by outbidding other countries and exerting international political influence.

> My next question is how many mask factories have we built in the US since we realized that was important?

We're so bad at building.

I'd add to your question with: how custom-built do factories need to be to be efficient?

I expect chip fab is very tailored, but can a t-shirt factory quickly pivot to surgical masks? How about N-95? Can a vacuum cleaner factory start building ventilators? How about HEPA filters? mRNA vaccines?

What are critical things we may need to ramp up rapidly, not just for pandemics, but for the next disaster, and can they be done without building dedicated factories just for those items?

Humans are fabulously general-purpose, so I'd imagine that the more automated the process, the more expensive it is to switch. This would put high-wage countries like the US at a disadvantage with regard to flexible domestic manufacturing. But I've no expertise in this area at all.

These are questions and guesses, not questions and answers. :-)


The machines to make mask material are very different from other textile machines.

They are expensive, and take a long time to build.

https://www.oerlikon.com/polymer-processing/en/solutions-tec...


Thing is, you are assuming that’s what is important.

Look at the billions the US spends on military, all of it domestic. In US, the masks are simply less important than all that, as is healthcare in general.


> My next question is how many mask factories have we built in the US since we realized that was important?

It's not important. Masks don't effectively stop the transmission of SARS2 and N95 masks are only highly effective against something as infectious as SARS2 Delta if you wear them precisely as intended (and to go with that you need to take many other precautions that the mass population is never going to take with great precision on a day to day basis). Even the vaccines don't stop Delta effectively, which you can plainly see from the large outbreaks in highly vaccinated nations.

> My question is why did it take this long?

Because the US Government spends nearly all of its effort on dumb shit, like shuffling papers, playing at admin, processing lobbyist appointments, screeching out empty promises every 2-4 years, going on talk shows and managing PR, managing a globalist superpower clown show, projecting power to every corner of the globe for no great reason other than to fulfill the powerlust of those in control.

They're supposed to be running our government for our people. They're simultaneously running the equivalent of one of the world's largest governments outside of our borders. Think about that for a moment. Do I think the people in power can run the US Government, domestically, effectively? Hell no. Do I think they can then simultaneously run a gigantic other foreign system - military bases, personnel, war, huge embassies, global trade, sticking their noses in every foreign political issue and at all times - on top of that? Triple hell no.


If China were to invade Taiwan, I'm guessing that all product flow from China/Taiwan to (most) other countries would stop.

If that happened, I'm curious what other manufacturing/product issues would impact other countries, and how long it would take them to adapt.

E.g., what tool manufacturing, raw materials mining, etc. are currently happening only in China and/or Taiwan?

And that's assuming that China didn't also sabotage industrial capabilities in other countries that depend on Chinese/Taiwanese manufacturing.


The U.S. supply chains would be completely wrecked. We're probably close to a point where in the event of a global war with China, we wouldn't be able to mobilize industry in the same way we did during WW2 simply because the industrial output fo yesteryear no longer exists in the U.S.

On the other hand, people forget the amount of IP (including semi design) that goes from the world to China/Taiwan to design products.

But that's what the whole point of free trade is supposed to be: increased interdependence, so that everyone has a gun to everyone else's head, and no one is incentivized to pull the trigger.


China can manage with loss of most or all access to American/NATO IP. America/NATO can't really manage without Chinese imports.

That's not accurate today. It might be different in 10 years, depending on how successful China is at becoming self-sufficient in semiconductor design and manufacturing.

Global companies didn't just outsource all the work to China. Even in the 90s, they were aware of the risk/reward, and specifically retained high-value, low-headcount/capital work outside of China.

As a result, China makes most of a device, but many core components are imported.

https://www.nasdaq.com/articles/huawei-revenue-slides-in-q3-...


We're not even talking about semiconductors. All of the basic goods you'd need to supply an army, or the industrial capacity for consumer goods that could be shifted to wartime production no longer exists. It's a problem the DoD has been talking about for years, but little action has happened.

Basic goods like food?

Imports, Total Estimated Value by Country, 2017 (most recent) [0]: Canada: 26,200.3, Mexico: 23,541.0, [...] China: 6,159.7

Or energy?

"In 2020, the United States exported about 8.51 MMb/d and imported about 7.86 MMb/d of petroleum1, making the United States a net annual petroleum exporter for the first time since at least 1949." [1]

Or arms?

Based and manufacturing in the US: Honeywell, Huntington Ingalls (ex: Northrup Grumman Shipbuilding), L-3, UTC, General Dynamics, Northrop Grumman, Raytheon, Boeing, and Lockheed Martin [2]

If you're talking a full manpower, total war scenario that requires mobilization of the entire country's non-military industrial base...?

I think there's a reason there haven't been any wars like that between nuclear armed powers. Ever.

Because by the time it gets to that stage now, someone has fired nuclear weapons if they have them.

[0] https://www.ers.usda.gov/data-products/us-food-imports/us-fo...

[1] https://www.eia.gov/energyexplained/oil-and-petroleum-produc...

[2] https://www.therichest.com/the-biggest/top-20-largest-arms-m...


They just steal the IP anyways

>> The U.S. supply chains would be completely wrecked.

China: Hey cool, we can destroy the US without engaging with them at all, we can just complete our takeover of Taiwan which we wanted to do anyway. We just need to make it disruptive enough to stop exports for a year or so - maybe a bit longer.


Yea just stop $500 billion in exports. That won't hurt the Chinese economy at all.

The best way to keep China from invading Taiwan is by upholding the One China Policy. With that policy in place, China has plenty of reasons to avoid a war. Unfortunately the US and many other parties are increasingly tearing this policy down, not seldomly without fully understanding why that policy was invented in the first place.

lol @ "it's Taiwan's responsibility to not be invaded."

You're like a middle school kid who justifies violence with "talk shit, get hit."

How much does China pay you to shill on here day after day?


Do you see what's happened/happening in Hong Kong?

Yes and you should read the book "The Other Side of the Story: A Secret War in Hong Kong" by Nury Vittachi to learn what the mainstream media wasn't telling you about the whole Hong Kong situation.

My Hong Kong friend who is pro China lives in an appartment complex where lots of yellow vests also live, and prior to the NSL every day he feared for his life, fearing that his neighbors would find out that he's pro-China and then do something to him. So much for the western coverage of the yellow vests being "pro-democracy, pro-rule of law, pro-free speech".


And yet your hypothetical friend is fine, but China has seized land again.

...seized land that belonged to them in the first place? It's the British that seized Hong Kong because the Qing refused to import opium.

agree

China has no reason to invade Taiwan, and plenty of reasons to avoid doing it. The whole thing is just another American wet dream.

Xi Jinping has been threatening invasion for years, and before that, decades of CCP threats.

https://www.bbc.com/news/world-asia-china-58854081


>he said unification in a "peaceful manner" was "most in line with the overall interest of the Chinese nation, including Taiwan compatriots".

Are you sure that’s the article you wanted to link to?


As far as American access to semiconductors goes, it doesn't matter if the takeover is peaceful or not. Chinese control over Taiwan could easily disrupt the supply chain through tariffs, export bans, backdoored hardware, etc etc.

Yes, China could do what the US is doing right now. So what?

Did you feel the same way about Hong Kong?

Seeing that Hong Kong was literally an NED-sponsored [1] color revolution, and that every country in the world had a national security law except for Hong Kong until last year, I'd say that the Hong Kong situation is not at all a case study for whether China would invade Taiwan. Since 2019 I've watched for a year how the western public and the protesters yelled that "it's going to be a Tiananmen 2.0" and "the tanks will soon roll in", only for the tanks to never have rolled in. In the mean time, the protesters who were called "pro-democracy" by western mainstream media were doing some very undemocratic things, such as assaulting fellow Hong Kongers for merely disagreeing with them[2].

Two Hong Kong youths have a Youtube channel on which they present views that don't align with mainstream western media. This interview is very telling: https://www.youtube.com/watch?v=zYK-cC9mw6c The guest on the show interviewed 100 protesters. What's interesting is that none of the protesters were able to explain what concrete change they want. The protesters merely blindly reigurated media talking points without being able to explain what's behind those talking points. None of them have actually read the NSL text.

Sound familiar? Yes: the "anti-voting fraud" Capitol rioters, who were hiding unnoble intentions behind a noble label. Many Hong Kong "protesters" were merely anti-China, not pro-democratic. They literally attacked people merely for speaking Mandarin. Ironically, one guy that was attacked was Taiwanese, not mainland Chinese.

Seeing how they literally burned innocent people[3], vandalized public infrastructure, vandalized mainland Chinese busineses merely for being mainland Chinese, made petrol bombs[4], etc. is it really reasonable to expect China to do absolutely nothing? Bringing back safety to the streets is what any government would do. After it was clear that they'd never send in tanks, people moved the goalpost and started demonizing the NSL but few people know what's actually in the NSL.

Demise of Hong Kong's rule of law are exaggerated. Even after the NSL, Hong Kong's judiciary remains fiercely independent — with evidence.[5]

[1] The NED is the regime change arm of the CIA. Despite the name, they are anti-democratic, having overthrown democratic regimes that refuse to align with the US. They publicly admit having sponsored the Hong Kong protests: https://www.ned.org/region/asia/hong-kong-china-2020/

[2] Videos: https://twitter.com/DanielDumbrill/status/116629626167206297...

[3] https://www.scmp.com/news/hong-kong/law-and-crime/article/30...

[4] https://twitter.com/SCMPHongKong/status/1195968903748317184

[5] https://www.scmp.com/comment/opinion/hong-kong/article/31480...


With many violent protestors being confirmed CCP agents, substantial amounts of social media manipulation, dismantling of the free press (eg, Apple Daily), imprisonment of journalists - this topic is so extremely muddied that none of your sources are going to convince anyone who is paying attention.

I think you are going to have to prove that "violent protesters are CCP agents".

You conveniently ignored his points about Apple Daily and journalist arrests.

> Seeing that Hong Kong was literally an NED-sponsored [...] color revolution ... such as assaulting fellow Hong Kongers for merely disagreeing with them[...].

Your comment is vastly disproportionate, and does not explain what happened.

Your comment suggests the following:

- Hong Kong’s protest was mainly due to sponsorship by the US, that is, it would not happen to this scale without US’s sponsorship (via NED).

- Hong Kong’s protest was mainly undemocratic (using your examples).

When discussing a social movement of this scale (millions went to the street, protests went on for half a year, with aftermath still affecting current events), we must judge whether your examples represent the majority or the extremely rare minority of what happened.

Just like the social movement led by Mahatma Gandhi, the movement led by Martin Luther King, or the recent BLM movement would not change much even if hundreds of protestors did things that go against the main thrust of the movements (which did happen).

To make this concrete: imagine blaming the BLM on CCP because they supported it [1], which is absurd [2].

We know that:

- The pro-democracy group had a landslide victory in the 2019 election (during the protests) [3].

- Many Hong Kongers left Hong Kong after the National Security Law [4][5]: a higher percentage of Hong Kongers left Hong Kong after the NSL than German left East Berlin after the split of Berlin.

Blaming the Hong Kong protests on CIA, NED, and the US does not explain the above events. Just like blaming the BLM on CCP does not explain the trend.

A much simpler explanation is that Hong Kongers do not like CCP rule, and they show it in the protests, in the election (when conducted fairly, without screening of candidates), and in their acts of leaving (when faced with the NSL and screening of candidates in future elections).

I suspect that the popularity rating of CCP, or of Kim’s regime in North Korea, would plummet if people know what is going on (that is, when the state censorship and propaganda are gone) and they are allowed to express themselves (for example, with a fair election). We just see this happened in Hong Kong.

[1]: https://www.japantimes.co.jp/opinion/2020/07/01/commentary/w...

[2]: https://www.nytimes.com/2020/09/18/technology/no-a-black-liv...

[3]: https://www.theguardian.com/world/2019/nov/24/hong-kong-resi...

[4]: https://news.yahoo.com/almost-90-000-people-left-040718900.h...

[5]: https://www.nytimes.com/2021/10/11/world/asia/hong-kong-popu...


You are right in criticizing me for not being proportionate enough. I did not take the time to make my post more nuanced. So let me clarify.

I do not believe that the HK protests were mainly due to US sponsorship (even if they were a contributing factor). My point was rather that it's unreasonable to expect a country to not respond to foreign sponsorship with security measures.

I do not believe that the HK protests were mainly undemocratic. However, the undemocratic and violent faction had effectively taken over, while the actually democratic faction failed in reigning this former faction in, so in my opinion trying to rhetorically separate them is a waste of time.

I do not agree with you that the CCP's popularity in the mainland is fake. There are multiple studies that show that their popularity is real.[1][2] The difference is between mainlanders and Hong Kongers is that many Hong Kongers don't have first-hand experience with how the mainland is governed and how the mainland has improved, and that the media regularly injects disinformation. Mainlanders know about Hong Kong, but they disagree. You can blame it on propaganda, but even mainlanders who have moved outside the mainland and who have been exposed to western media, don't tend to change their stance on Hong Kong.

[1] https://ash.harvard.edu/files/ash/files/final_policy_brief_7...

[2] https://www.washingtonpost.com/politics/2021/05/05/did-pande...

What bugs me most is that the criminals are getting away with a good reputation in the west, who wholly paints them as "pro-democracy freedom fighters".

It also bugs me that a lot (not sure how many, but a lot) of protesters don't know what sort of freedom they want. Check the Keybros interview I linked: why could none of the 100 protesters articulate what concrete freedoms they are missing? Why are the protesters mostly younger people who have never lived under British colonial times? There is something deeply disingenuous among a part of the protester movement. I hope they can recognize that economic trouble (e.g. cage houses) is one of the reasons, and I hope that Hong Kong will do something about that. If these protesters think that life in Britain is really better, then they should experience it firsthand and come to their own conclusion on whether their protest was right or not.

What I hope for is more reconciliation between mainland and Hong Kong. I even hope that Hong Kong will get more democracy, namely universal suffrage and a strong Common Law-based rule of law environment. However I want Hong Kongers to be honest about things, and I want more democracy/rule of law at the same time as a Hong Kong that is proud of being part of China, and at the same time where criminals aren't swept under a rug for political reasons. One Country Two Systems, not Two Country Two Systems.


> I do not agree with you that the CCP's popularity in the mainland is fake. There are multiple studies that show that their popularity is real.[...][...].

I am not questioning CCP’s current popularity in Mainland (or Kim’s current popularity in North Korea), given their current control (including in education and media, both propaganda and censorship). When the population were brought up that way, even when presented with opposing view points their thinking is skewed (compared with many other democratic countries), and considers things as conflicts of super powers or humiliation by foreigners. Such thinking frequently persists and affects their actions even after they move outside of China (censor Hong Kong and Xinjiang in FANNG [1]), even in face of opposing evidence (linking Xinjiang to CCP [2]). Many of the Chinese population are effectively exporting their values to the world [3].

For a concrete example, even when foreign sponsorship does not pose material threats, some Chinese would suggest disproportionate responses (such as to consider draconian security measures and the National Security Law, resulting in arrests of the majority of democratic candidates [4]). This is disproportionate because UK and other European nations would not take comparable measures against comparable foreign sponsorship (by CCP). Just like in the US, Texas Secession could be discussed and supported without similar consequences to the NSL in Hong Kong [5]. This shows the resilience of their political systems, and by contrast CCP is overreacting.

Gaining popularity this way may be fine for CCP (or for Kim) for now, but when the system becomes unstable (due to declining economy and shrinking population), things may start to crack. The problem is not that it is fake, the problem is that the system (by censoring dissent and pointing fingers at others inconsistently) is unsustainable. And at least, this is why CCP and Kim do not currently have as high popularity in Hong Kong, Taiwan, South Korea, Japan, Sweden, Czech, Germany, France, or elsewhere.

[1]: https://news.ycombinator.com/item?id=27649830

[2]: https://news.ycombinator.com/item?id=29384161

[3]: https://news.ycombinator.com/item?id=21194083

[4]: https://apnews.com/article/legislature-primary-elections-dem...

[5]: https://www.texasmonthly.com/news-politics/obama-administrat...

> It also bugs me that a lot (not sure how many, but a lot) of protesters _don't know what sort of freedom they want_.

Many Hong Kongers want CCP to leave them alone, which was the implicit contract behind the 1997 hand over of Hong Kong. To be more specific, Hong Kongers want to run the city their way (including have their elected representatives not being arrested [4]), and not having laws inserted into their mini-constitution (Basic Law) against their will, to respect their differences—more or less the way Hong Kong was run before 2019. Incidentally, I think that’s what most Tibetans originally wanted, and what most Taiwanese currently want (against unification), too.

Wanting CCP to leave them alone was the main theme behind the 5 demands in the 2019 protest. What they don’t know is how to achieve it, when faced with CCP, the PLA and the massive Chinese population.

I am less worried about them not knowing exactly what or how to achieve the freedom they want, just like the population of Singapore, Taiwan, Japan, or South Korea might not know exactly what they wanted during their democratization (I would not be surprised by a Keybros interview when they democratized. BTW, if you want to know what the protestors think outside of cameras or Keybros, check out https://www.thestandnews.com/). Ignoring the selection bias, not everyone is as articulate, but this does not stop the trend of democratization. They democratized by trial and error 摸著石頭過河, and things work out fine for them. Same for Hong Kong.

> why could none of the 100 protesters articulate what concrete freedoms they are missing?

How did the Keybros interview avoid the selection bias?

> Why are the protesters mostly younger people who have never lived under British colonial times?

The younger ones have less burden and could go to the streets, the older ones with more burden supported them. See the election result for the proof [6].

> There is something deeply disingenuous among a part of the protester movement.

Hong Kongers want to say No to CCP, but this could not be said out loud with the PLA standing by. This is the disingenuous part. (Just like many protests in China would be prefaced with a support of the CCP, however contradictory and disingenuous sometimes it could get. Also, many Taiwanese wanted to say so, but again need to consider the threat of the PLA.) This is why many Hong Kongers left when they still could [6], they vote with their feet, if they could not vote with their hands.

[6]: https://news.ycombinator.com/item?id=29346205

> One Country Two Systems, not Two Country Two Systems.

CCP said that first to Tibet in 1951, then to Hong Kong in 1984. Seeing what happened, now in 2021 even the Nationalist party KMT in Taiwan is rejecting One Country Two Systems. Sometimes we need to learn from history.


This will cause WWIII so yes we will need chips in that scenario. Let’s hope that doesn’t happen.

Maybe it won't cause WWIII

Samsung could build the plant in South Korea?

Same situation. Most of South Koreas infrastructure / urban centers would be decimated in an 'all out' conflict with North Korea.

[1] https://www.forbes.com/sites/michaelpeck/2020/08/11/north-ko...


In particular, all of Seoul is in artillery range of the north

Rusty dusty equipment that probably won't even fire.. that offensive line is hardly a major threat

NK holds regular military exercises. And 152mm guns are tried and true; they aren't exactly the most finicky pieces of equipment.

Here's a massive artillery drill they held just a few days ago.

https://www.bloomberg.com/news/articles/2021-11-07/north-kor...


It doesn't matter if it is a 100 year old artillery piece, they aren't exactly complicated, just large breach loaded guns. The only thing modern guns and artillery have over old guns is accuracy, which doesn't really matter when you have enough of them firing down range at a massive target.

What's to stop China moving in with theirs? I mean after all they're friends and that's what happened during the lasy war...

> the "Let's build a Fab in the USA" train

We have fabs in the USA right now. To speak of Austin alone, Samsung's own S2 has been here since the late 90s, and there are a handful of others here, too.


Globalization led to an erosion of US wages and the hollowing out of our middle class.

If we buy local, it creates domestic jobs. Wages go up too.

We need to onshore more.


That's so many assumptions you might as well put on a robe and special hat and become a fortune teller.

The effective erosion itself is a policy choice, there is this thing called 'minimum wage' which you can use to make sure that there is more than just 'warm bodies' as the lowest common denominator in job requirements, and as a result everything above that has to be 'upgraded' to a higher standard as well, otherwise people aren't willing to do the work (just like is happening now). The examples are all around you.

There is no onshoring of talent or natural resources since they are not 'constructed' the same way and importing those are against the current ideas/fears of immigration and relying on third parties for goods.

In other words: globalisation isn't a choice, it's a side-effect from the needs of humanity. Reverting to a more primitive state that decreases those needs could be a path, but I doubt anyone would choose that willingly.

Besides those things, there is this concept of ideals or principles which are almost as important and it doesn't require any extreme form of meddling either. If people really care about something (say, the well-being of other humans and therefore not really being happy with an authoritarian regime somewhere) then that can be incentive enough. And if it isn't then some protectionism won't be enough either.


If you think bringing manufacturing to the US is about jobs, then I have a bridge to sell you.

> We need to onshore more.

Simple solutions to overly complex problems is not the answer.

First, manufacturing output has steadily increased since the 90s. https://fred.stlouisfed.org/series/INDPRO

Second, the number of manufacturing jobs have steadily declined since the 90s. https://fred.stlouisfed.org/series/MANEMP

How is this possible? Automation. These plants aren't creating US jobs and increasing wages, they're simply creating more wealth for corporations which they would have done anyway overseas. And since Samsung is a Korean company, the wealth is being created in Korea, not the US.


> they're simply creating more wealth for corporations which they would have done anyway overseas

They're also resolving some pretty serious national security concerns, so...


Protectionism makes everyone worse off.

>> Protectionism makes everyone worse off.

Yeah I mean look at how China has grown the last 25 years with their fully open markets. </sarcasm>

Every approach has pluses and minuses. Putting your own first is common and healthy. The US has been failing to do this for some time now, and the economy is pinned in a corner with no easy way out.


China is now much more open then they were for a long time and as it happens their growth really started when they did that. The are 1.4 billion people who have embraced mostly 'typical capitalism' that is comparable to most countries in the world do. Its not like they are the first to have the idea of tarrifs or helping strategic industries. China has continental scale and 100s of millions of people living on the coast with billions of people in close distance to export to. Of course they are gone do well, the internal market they generate alone is absurdly gigantic.

Just like the US in the late 1800 century, they also had some protectionism. But it crazy to suggest that is the main reason why they were successful.

Nobody says government should make important technology investment, have some strategic supply chain and so on. However just blanked Trump style protectionism across random industries is not really the solution to anything.

> Putting your own first is common and healthy.

This is a myth of what protectionism achieves. You protect some sector but potentially damaging other sectors considerably. The US protects Flordia suger farmers and as a result everybody else pays more for sugar.

Most of the time its groups that can lobby effectively that get tariffs and those groups then profit against the benefit of everybody else.


The pushback from many whenever it’s suggested that the US should undergo a general trend to bring back manufacturing jobs to America always fascinates me. I always wonder what’s going through their heads when they prefer that things continue to be made on the other side of the world in a sweatshop paying their workers cents on the dollar to make cheap, soulless, throwaway goods in countries with terrible environmental laws, where they’ll then be shipped using a ton of fuel to get those goods across the world to us.

Whatever happened to taking pride in what we make and buy? Quality over quantity? Paying more for something better? Looking out for your neighbors and community by supporting their gainful employment? Independence from countries that literally hate us?

It wasn’t long ago at all that we were making most of the stuff we bought. It’s time to go back to that. Call out every company you see that doesn’t make their products in the USA and ask when they’re going to stop selling out their own neighbors just for a buck.


It is bordering on racist to suggest that because things are US produced they will be Quality over Quantity. What you actually need to win is Quality and Quantity. The same arrogance the US manufactures had about Japan and got their as kicked.

At seems like you have some delusional fantasy that the world could run on locally made artisanal goods and that those goods then magically have more 'soul' (whatever that is) and would be better for the environment. And even if that was the case, these products would still depend on international supply chain.

Its equally false to suggest that everything outside of the US is made in sweatshops. China doesn't dominate because of sweatshops, but because going down stream, mining, refining. These are complex operations, requiring lots of educated engineering, capital and vision. Look at a company like CATL that is powering much of the worlds EVs. This is a company that exploded in growth in a highly complex industry.

> Paying more for something better?

So like car from Japan, Europe and soon China?

Where are you getting those high quality electronics, batteries, refined rare earths and so on? Please tell me the US companies who produce all these amazing things that are so much better then those from the international competition

> It wasn’t long ago at all that we were making most of the stuff we bought.

Who is 'we'? The US? Your family? Are you advocating to go back to per modernism where every woman was doing 40h of weaving for the family to have cloth?

The US never produced everything it needed. In modern history international supply chains are a thing.

And this was most true after the US literally bombed the shit out of Europe and Japan. And China was destroyed by Japan. Soviets and China turned communist. Maybe do that again and you can again do that? Even then it was the US and the British empire.

> Call out every company you see that doesn’t make their products in the USA and ask when they’re going to stop selling out their own neighbors just for a buck.

You can't fix problem by calling people out. If you force companies to make 'everything local' then they will simply no longer be internationally competitive.

If you seriously think to just cut of the US from the rest of the world would lead to a higher standard of living you are sorely mistaken.

Your level of understanding of the connection between international trade and living standards seems to be totally lacking. Your suggestion to just 'guilt' (or assuming with fines or whatever) every company into buying all local is a surefire way to completely destroy the economy and the competitiveness of the US.

And its simply anti-democratic. People might be willing to pay 1% more if something is marketed as locally produced. But most of the time people don't actually care. And they never cared. People care about themselves and their families and their standard of live. If Japan makes better cars, they will not simply buy American because its made in Detroit. The same goes for everything else.

If you want to improve standards of living, just cutting yourself of from international trade is literally the worst thing to do.


> It is bordering on racist

I stopped reading right there, in the future please hesitate to use self-defeating arguments like this in your counter arguments.


Hard to hear the truth about your own ignorance.

Opening up trade doesn’t necessarily make everyone better off. If makes everyone better off overall, at least to first order, but if you have national security requirements for needing domestic capacity anyway, that may make a bigger difference.

And you can make trade better for everyone period if you redistribute some of the gains of trade to those that it makes worse off. But… we don’t always do that…


How?

It's not protectionism. It's anti-fragility. It's having strong domestic capabilities to meet our every need and not suffer when the world is uncertain.

It's having jobs and prospects for our future generations. The hope Millennials and Gen Z have is dim, and I can point my finger at a huge contributing cause.


False. Exhibit A: China from 1980-Present. QED.

China's growth has been mostly due to the liberalization of their markets. You're proving me right.

There’s huge difference between protectionism and investing locally.

Does it? Early covid we didn't even have enough PPE. Countries that produced PPE had it.

Yes, but why? Let me answer, because slave labor is cheaper.

not cheaper than robot labour

Hollowing out the industrial base leaves us even worse off.

It only leaves lower tiered hierarchies worse off and only in the short term. Extracting value that way is never a long-term plan since the needs will either be fulfilled or the extraction will be complete and no more value can be created.

If people don't want to buy a product because it is inferior in some way, the solution isn't to ban superior products but to upgrade those inferior products to be superior again. And inferiority/superiority doesn't just mean quality, it can also mean features and the price-quality balance in itself. At the same time, the rest of the world doesn't "go away" if you construct an island within your borders that nobody is allowed to join, so at best it's going to end up in some sort of useless industrial-cold-war.


> It only leaves lower tiered hierarchies worse off and only in the short term.

This is the refrain of those who do not understand reflexivity, and third-order analysis.

This analysis is correct to the first and even second-order effects, but completely misses the importance of slipstreaming floor expertise / experience into design and engineering iterating, or even holistic support. This kind of economic caste system thinking just as surely shackles the US as the Indian social caste system shackles that nation with enormous untapped potential.

I routinely make LinkedIn friends and actively solicit input from the operations people in all of my client organizations, and this has repaid me many times over. Where others within their very own organizations, much less outside consultants, snub these "lower tiered hierarchies", shedding light upon their experiences and expertise has taught me many valuable lessons about software design, debugging, implementation, and product management, among many other areas.

I learned to do this from reading about the nuts and bolts of how Japanese automakers applied Deming's quality principles. IMHO the buried lede in those stories wasn't all the charts and reports artifacts (similar to the Agile artifacts we use in our industry), but to my mind it was the attitudinal changes that flew in the face of common human hierarchical predispositions. Because it is a massive organizational cultural shift often enmeshed in wider prevailing ethnographic contexts, it is one of the most difficult leadership-led transformations to conduct.

One can get away with the hollowing out effect in the beginning, and if there is sufficient "mass" of value to strip in the industry to mask the effect, it can even be carried out through a couple more product and technology cycles to second-order ramifications as various affiliated supply chains are also stripped out. But once it reaches third-order ramifications often loftily projected as, "while we lead the toiling-for-peanuts offshore masses with our brilliant sales/marketing/product/accounting insights, the supercharged stock bonuses will just roll on in!", it is when reality slaps them in the face and they find out that those same "junior", "lower tiered hierarchies" toss aside the "onshore leadership" to cut out the middlemen to the customers.

After one is never too proud nor too lofty to lead from the front, be happily willing to (if even for a moment) do anything one would ask their direct reports or even their recursive direct reports to do, so to speak "sleep on the factory floor" if need be, a torrent of insights that would normally do not occur during design and engineering stages become available.


I don’t know if it will bring down production costs, but it may reduce logistical shortages and bring down chip prices.

It’ll reduce geopolitical strategic risk.

is there a labor demand for electronic assembly? we can barely get workers for decently tipped restaurants, and we now going to have 1000s lining up to twist a screw for 8 hours straight?


> This is about the war with china.

> It's all an attempt to prevent war with China.

So the Chinese have UFOs and we have a war with China but we are moving chip production to all these countries you mentioned to prevent the ongoing war with China?

I’d like to give you the benefit of the doubt here though. I think if you were able to be a little more clear I’d at least understand what you intend to say, but I find your comments here a bit confusing.

> And nobody is talking about it? This is completely related to the war with china.

Who isn’t talking about it? Assuming you’re talking about poor relations (“Cold War”) and/or loving chip production. It seems to me to be pretty common knowledge and is widely reported in global media sources and the Internet. Don’t the links you provide demonstrate that?


We’re moving fabs to the West to defuse a potential hostage situation where China strangles the global supply of semiconductors.

The UFO releases are either about telling China that we are aware of and can track their surveillance assets, showing China our new highly advanced classified assets, or else demonstrating the superiority of our current Air Force assets.

Both activities serve to dissuade the Chinese military from taking aggressive actions.


Except that 1. It’s the west (in particular the US) that is strangling the semiconductor supplies in order to hamper Chinese industry, and 2. China is completely uninterested in military aggression, as opposed to US, which just finished a war and is probably looking for another one.

This is just wrong.

The White House recently identified semiconductors as an essential product critical to US national security [1]. These actions by the US and allies to onshore fabs is part of the adjustment to defuse a supply chain risk caused by potential future actions by China.

During the Cold War, both sides used strategic release of classified information on capabilities to notify the war planners on the other side in order to prevent miscalculations that could lead to a nuclear exchange. This is common knowledge and a good tactic to tamper escalation. It would be extremely irresponsible to surprise a nuclear peer with new capabilities.

[1] https://www.whitehouse.gov/briefing-room/statements-releases...


That's not what this is about. The US is sanctioning SMIC. It forced TSMC to release customer data of mainland Chinese customers, presumably with the goal of checking which more Chinese companies to sanction. The US's goal is to prevent China from building a self-sufficient semiconductor industry. Or conversely: to keep China dependent on the US so that the US can sanction China at any time.

Yes, SMIC looks like one aspect of a global reorganization of semiconductor manufacturing.

I’m not sure I buy your reasoning though. I don’t understand how effective blocking the use of US technology by SMIC will be given the propensity for IP theft, so I don’t see how it furthers the goal of preventing China from becoming self-sufficient. Nor would I really agree with the assumption that they are not already self-sufficient.


US rethoric says that they are concerned about IP theft, unfair economic practices, etc. But if you look at the sanctions then you see that they target pretty much only R&D-heavy sectors. That should be telling.

The semiconductor industry is extremely R&D heavy. You can't just steal your way into success. It is also heavily dependend on a global network of specialized suppliers and is one of the most complex supply chains on earth.

Even Chinese media says that they are not yet self sufficient. They predict that SMIC will become able to produce 28nm and above using only domestic inputs by the end of 2022, but that's just ability and doesn't factor in manufacturing scale to handle the corresponding demand.

Making 14nm fully domestic takes a bit longer -- maybe 2 years more. Chinese media and experts say that if they scale up 14nm and above, then that satisfies 70% of the demand.

To go further than about 7nm you need EUV lithography, which only has 1 supplier in the world, and the US is blocking that. How long it takes for China to independently develop EUV is unknown. Universities have made small breakthroughs in subfields here and there but they're still far away from the full solution.


There’s a difference between self-sufficiency and state of the art. China can be self-sufficient without having cutting edge technology.

For most military applications, 28nm process is comfortably sufficient - USAF 5th generation aircraft were designed and flying before 28nm.

I’m not convinced that there is any real need right now for <14nm process besides striving for market share. Market share which is itself reorganizing around new ISAs. EUV is a (amazingly clever) Rube-Goldberg machine.


> China is completely uninterested in military aggression

Except they increasingly fly their fighter jets in Taiwan's airspace. And regularly engage in deadly border disputes with India. And continue to conduct cyberwarfare.

https://www.bbc.com/news/world-asia-58794094

https://en.wikipedia.org/wiki/2020%E2%80%932021_China%E2%80%...

https://en.wikipedia.org/wiki/Cyberwarfare_by_China


Except that Taiwan's ADIZ is not its sovereign airspace, but a unilaterally decided zone that covers large parts of mainland China.

https://independentaustralia.net/politics/politics-display/l...

Except that the border dispute with India was caused by the British, who randomly drew some borders somewhere that parties then later disagree with. India also has border disputes with non-China neighbors. China already tried to solve this diplomatically, for example by trading areas that India claims but are defacto China-controlled and vice versa, but India refused all deals so far because they want to claim in all.

Do note that China (PRC) has already given up land in the past as part of diplomatic deals. Check Vladivostok. Check the whole of Mongolia.

Western coverage of China’s border conflicts are extremely distorted. In the ASEAN summit, southeast asian nations are upgrading strategic ties with China.

https://www.france24.com/en/video/20211028-asean-summit-sout...

Many countries are worried about AUKUS. Malaysia planed China consultations as anxiety simmers over defence pact.

https://amp.scmp.com/week-asia/economics/article/3149713/mal...


>So the Chinese have UFOs and we have a war with China but we are moving chip production to all these countries you mentioned to prevent the ongoing war with China?

Is it China's UFOs? I don't think so.

https://www.cnbc.com/2020/02/28/elon-musk-says-the-fighter-j...

Elon is 100% correct the fighter jet era is over. Did you know Elon was on Trump's tech council? https://www.engadget.com/2017-06-01-elon-musk-leaves-trump-c...

>I’d like to give you the benefit of the doubt here though. I think if you were able to be a little more clear I’d at least understand what you intend to say, but I find your comments here a bit confusing.

The blow your mind factor is hard to address because of the roswell factor. I 100% don't believe those UFOs are extraterrestrials.

>Who isn’t talking about it? Assuming you’re talking about poor relations (“Cold War”) and/or loving chip production. It seems to me to be pretty common knowledge and is widely reported in global media sources and the Internet. Don’t the links you provide demonstrate that?

I guess I didn't make that clear. Obviously it's being talked about.

I guess step back for a second. The war with china is obvious. I never even talked about cyber aspect of the war, its not relevant exactly. A nuclear bomb is going to do more in the war than some BGP route stealing tech. The irony of it all, it's China's capitalist zones that are going to be robbed for 'common prosperity' at the same time that the cold war kills those zones like Evergrande situation.


How does research on poorly understood atmospheric phenomena relate to war with China?

If there's danger of a peasant revolt, war with the US would be a pretty effective way to prevent it.

French Economist article archive: https://archive.md/fcTkA ("brigades, or a division" sure doesn't sound like WWII)

Grauniad AUKUS article archive: https://archive.md/LFsoo


OK, a lot to unpack.

Lots of militaries prepared for conflict. OK.

What do the pentagon UFO releases have to do with war with China? You say it's completely related, would you elaborate?

And a peasant revolt in China? That's a big statement and I'd like to know more about your reasoning.


Best standard of living for Chinese "peasants" in history. Not sure where this revolt comes from

The easiest (and probably the only) way to prevent a war with China is to convince the US to not start it.

>The easiest (and probably the only) way to prevent a war with China is to convince the US to not start it.

Yes, if there's one thing we should have learned from WWII, it's that capitulating to land-grabs results in the bad-actor stopping. If we just let them take over the entirety of the South China Sea, Taiwan, and maybe just half of the African nations they've been lending money to who will never be able to repay their debts, they'll definitely be satisfied and stop. Korea, Japan, and India have nothing to worry about. Nothing at all...


Until you realize that you are fine with the same stuff being done by US.

Your response has literally 0 substance. Ignoring the fact that whataboutism is a tired and lazy rebuttal, remind me which country has the US has annexed in the last decade, or whatever you’re trying to claim. I haven’t heard Canada or Mexico complain that they’re concerned we’re going to invade.

Remind me what country has China annexed in the last decade?

Your arguments were about eg Africa, which is obviously similar to the Marshall Plan.


> Remind me what country has China annexed in the last decade?

They’ve been annexing parts of Nepal for the last decade. They are attempting to claim parts of the South China Sea that belong to multiple neighbors.

Violating the agreement they made with the UK is essentially annexing Hong Kong.

The Marshall Plan was absolutely nothing like what China is doing and to claim so is utter ignorance of what is happening whether intentional or not. Given your responses so far I’ll assume it’s intentional.

You still have not provided a single example of what you’re accusing the US of.


>The easiest (and probably the only) way to prevent a war with China is to convince the US to not start it.

Noam Chomsky recently wrote similar to this. The military threat is really against china.

The point of a cold war is to prevent a real war. So when you hear about dick waving of the US military and allies. IT's about warning China from doing anything.


Who do you think is the aggressor? China or the US?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: