4 ms·

It is crazy to me, that this investment will be all in Seoul. It is not very hard to make the case that Seoul is overly concentrated for both politics and the

It is crazy to me, that this investment will be all in Seoul. It is not very hard to make the case that Seoul is overly concentrated for both politics and the economy at the expense of other areas of South Korea. And as far as security balwark goes, Seoul is only 40km from the DMZ.

Barrin92 4y ago
Chip supply chains are some of the most complex things humans have ever built and the amount of skilled work you need is crazy, so it really doesn't make much sense to put it anywhere else. In general also many East-Asian cities don't really suffer from dysfunction and lack of development so agglomeration really has a lot of upsides.
- bobthepanda 4y ago
  The cost of housing in Seoul is pretty dysfunctional, as is the general commuting crush: https://m.koreatimes.co.kr/pages/article.amp.asp?newsIdx=339661 https://m.koreatimes.co.kr/pages/article.amp.asp?newsIdx=339...
- bsaul 4y ago
  do you have any good resources i could read on the subject of chip supply chain ? I know absolutely nothing about it
  - jiggawatts 4y ago
    I wish I had ready-to-go URLs I could reference, but the complexity is just stunning. For example, the purity of the feed chemicals has been a key limitation, and steady improvements have been made by suppliers over decades. There are orgs that specialise in things such as shipping ultra pure chemicals in specially cleaned teflon containers. Just trucking and piping this stuff around must have had millions (billions?) of dollars of research sunk into it. You don't get to 99.9999% or better purity while pumping stuff into a dirty old diesel truck's rusty tank! The "tools" themselves are an extra level of crazy. For one, the ASML x-ray ("EUV") laser lithography machines are just an unbelievable piece of precision engineering. If twenty years ago you had told me that we'd have high-power x-ray lasers in an industrial setting with nanometer focusing precision I would have laughed in your face, or assumed that it's a function of a particle accelerator the size of a decent town. I remember seeing an add for a new-generation tool (machine) designed for the permanent vacuum part of the production line. They were proudly showing off their key improvement, which was that its MTBF was something like 7 hours. This was amazing apparently, because both them and their competition couldn't do much better than 40 minutes previously! Think about how demanding the requirements must be that a 40 minute MTBF was state of the art and totally acceptable in large-scale manufacturing, and raising that to "still less than a day" is worthy of advertising in an industry magazine!
  - rdl 4y ago
    Asianometry youtube/substack/etc.
LatteLazy 4y ago
Leaving aside the DMZ/NK Risk... Cities work. Bigger cities work better than smaller cities. They're cheaper, cleaner, greener, more productive, with better health and social outcomes and more access to education etc etc etc. That's why basically any large project basically NEED to be within X minutes of a major city. The saving in land cost setting it up in a rural area is nothing compared to having no staff, no services, no infrastructure, local political problems etc.
- AdamN 4y ago
  For fabs it seems like just a bit outside of a city is the optimum location. Chandler (Phoenix) for Intel's new fab and outside of Portland for the other US ones. I expect it's similar in Taiwan and South Korea. Side note but isn't water a critical input for a fab? Arizona's water problems are going to continue to get worse so this seems like a bad move by Intel.
  - LatteLazy 4y ago
    Yeah, it makes sense to put it not in the centre but within (say) 1h of the centre. Then you can recruit from an 8 figure workforce including multiple universities, offer a huge list of places to live and amenities to workers, meet reliability with everyone from regulators to ambassadors to investors to lawyers, have gold standard infrastructure, etc. I'm never sure when it comes to water. I think a lot of places in the west chose to have shortages by refusing to build infrastructure and wasting supply on subsidizing things like (export orientated, unnecessarily high water use forms of) agriculture. But I am no expert...
- bobthepanda 4y ago
  There is more than one city in South Korea. Over half of South Koreans live in the Seoul Capital Area. That’s a much higher proportion than countries other people often accuse of being overcentralized and neglecting the regions, like the UK with London or France with Paris. And to some degree what you describe is a self fulfilling prophecy; big cities hollowing out the smaller ones, shrinking their economies and leaving their governments in deficit and the services worse, rinse and repeat.
rickdeckard 4y ago
The location will actually be Gyeonggi-do, south-east of Seoul. This area is close to the existing semiconductor regions of S.Korea, such as Pyeongtaek, Giheung, etc. So it's quite an organic decision to select such a location, as also all the suppliers and customers are already located in this region. Relocating to an entirely different region would just delay the process and actually risk to collapse the economy for quite mundane reasons. Other Industries have their center of weight located more south, like Automotive and Energy. One "mundane reason" to imagine the complexity: People now working in semiconductor sector all live around Seoul. Apartments in Korea are not rented but predominantly purchased on loan. Just announcing to pull a whole industry further away from Seoul would immediately drive up housing prices of the destination and collapse the housing prices in Seoul. So people selling to relocate could not pay off their loans with that money and could not acquire at the destination due to huge debt. Seoul real-estate market could collapse, and since most of it is owned on loan, it would at least immobilize the whole Seoul region (as people could not relocate without risking bancrupcy), at worst cause banks to no longer accept the apartments as sufficient security for the loan.
- kijin 4y ago
  I think you're grossly exaggerating the real estate implications. There are 25 million people in the capital region, and only a small fraction of them work in the semiconductor industry. Relocating 200-300K people to a new industrial center might ease the pressure on housing a little bit, but there are millions more waiting to move in. Moreover, Seoulites really, seriously, desperately don't want to move out. The government has tried many times to create new cities away from Seoul with stable public-sector jobs; but a lot of people who work in Sejong, for example, just keep living in Seoul and make the 250km round trip every day. The 300K who actually moved to Sejong didn't make a dent in housing prices anywhere else, either.
- pigsty 4y ago
  > Just announcing to pull a whole industry further away from Seoul would immediately drive up housing prices of the destination and collapse the housing prices in Seoul. That’s a win for the average person in Seoul. Housing prices are destroying this generation of young people.
seoulmetro 4y ago
They have tried building hubs further out of Seoul, but it just doesn't work. Why would people leave and tie a billion liabilities to their workplace which is not even easy to maintain in Seoul itself? They want workers and they want Seoul. Workers want Seoul and they don't want to rely on their employers. This is always going to continue the fueling of Seoul border growth.
- bobthepanda 4y ago
  But workers want Seoul because already all the eggs are in the Seoul basket. Universities? Seoul. Jobs? Seoul. Services? Seoul. The country has been overconcentrated in Seoul since at least the beginning of the dictatorship. Breaking the cycle will not happen overnight, and you don’t even need to build new cities, just invest in the other ones. (Sejong was probably a bad idea.)
  - seoulmetro 4y ago
    It's the same in every country. It's just that Korea started so late that there were no other cities to build from. Busan isn't a great city to grow.

Barrin92 4y ago

Chip supply chains are some of the most complex things humans have ever built and the amount of skilled work you need is crazy, so it really doesn't make much sense to put it anywhere else. In general also many East-Asian cities don't really suffer from dysfunction and lack of development so agglomeration really has a lot of upsides.

bobthepanda 4y ago
The cost of housing in Seoul is pretty dysfunctional, as is the general commuting crush: https://m.koreatimes.co.kr/pages/article.amp.asp?newsIdx=339661 https://m.koreatimes.co.kr/pages/article.amp.asp?newsIdx=339...
bsaul 4y ago
do you have any good resources i could read on the subject of chip supply chain ? I know absolutely nothing about it
- jiggawatts 4y ago
  I wish I had ready-to-go URLs I could reference, but the complexity is just stunning. For example, the purity of the feed chemicals has been a key limitation, and steady improvements have been made by suppliers over decades. There are orgs that specialise in things such as shipping ultra pure chemicals in specially cleaned teflon containers. Just trucking and piping this stuff around must have had millions (billions?) of dollars of research sunk into it. You don't get to 99.9999% or better purity while pumping stuff into a dirty old diesel truck's rusty tank! The "tools" themselves are an extra level of crazy. For one, the ASML x-ray ("EUV") laser lithography machines are just an unbelievable piece of precision engineering. If twenty years ago you had told me that we'd have high-power x-ray lasers in an industrial setting with nanometer focusing precision I would have laughed in your face, or assumed that it's a function of a particle accelerator the size of a decent town. I remember seeing an add for a new-generation tool (machine) designed for the permanent vacuum part of the production line. They were proudly showing off their key improvement, which was that its MTBF was something like 7 hours. This was amazing apparently, because both them and their competition couldn't do much better than 40 minutes previously! Think about how demanding the requirements must be that a 40 minute MTBF was state of the art and totally acceptable in large-scale manufacturing, and raising that to "still less than a day" is worthy of advertising in an industry magazine!
- rdl 4y ago
  Asianometry youtube/substack/etc.

LatteLazy 4y ago

Leaving aside the DMZ/NK Risk... Cities work. Bigger cities work better than smaller cities. They're cheaper, cleaner, greener, more productive, with better health and social outcomes and more access to education etc etc etc. That's why basically any large project basically NEED to be within X minutes of a major city. The saving in land cost setting it up in a rural area is nothing compared to having no staff, no services, no infrastructure, local political problems etc.

AdamN 4y ago
For fabs it seems like just a bit outside of a city is the optimum location. Chandler (Phoenix) for Intel's new fab and outside of Portland for the other US ones. I expect it's similar in Taiwan and South Korea. Side note but isn't water a critical input for a fab? Arizona's water problems are going to continue to get worse so this seems like a bad move by Intel.
- LatteLazy 4y ago
  Yeah, it makes sense to put it not in the centre but within (say) 1h of the centre. Then you can recruit from an 8 figure workforce including multiple universities, offer a huge list of places to live and amenities to workers, meet reliability with everyone from regulators to ambassadors to investors to lawyers, have gold standard infrastructure, etc. I'm never sure when it comes to water. I think a lot of places in the west chose to have shortages by refusing to build infrastructure and wasting supply on subsidizing things like (export orientated, unnecessarily high water use forms of) agriculture. But I am no expert...
bobthepanda 4y ago
There is more than one city in South Korea. Over half of South Koreans live in the Seoul Capital Area. That’s a much higher proportion than countries other people often accuse of being overcentralized and neglecting the regions, like the UK with London or France with Paris. And to some degree what you describe is a self fulfilling prophecy; big cities hollowing out the smaller ones, shrinking their economies and leaving their governments in deficit and the services worse, rinse and repeat.

rickdeckard 4y ago

The location will actually be Gyeonggi-do, south-east of Seoul. This area is close to the existing semiconductor regions of S.Korea, such as Pyeongtaek, Giheung, etc. So it's quite an organic decision to select such a location, as also all the suppliers and customers are already located in this region. Relocating to an entirely different region would just delay the process and actually risk to collapse the economy for quite mundane reasons. Other Industries have their center of weight located more south, like Automotive and Energy. One "mundane reason" to imagine the complexity: People now working in semiconductor sector all live around Seoul. Apartments in Korea are not rented but predominantly purchased on loan. Just announcing to pull a whole industry further away from Seoul would immediately drive up housing prices of the destination and collapse the housing prices in Seoul. So people selling to relocate could not pay off their loans with that money and could not acquire at the destination due to huge debt. Seoul real-estate market could collapse, and since most of it is owned on loan, it would at least immobilize the whole Seoul region (as people could not relocate without risking bancrupcy), at worst cause banks to no longer accept the apartments as sufficient security for the loan.

kijin 4y ago
I think you're grossly exaggerating the real estate implications. There are 25 million people in the capital region, and only a small fraction of them work in the semiconductor industry. Relocating 200-300K people to a new industrial center might ease the pressure on housing a little bit, but there are millions more waiting to move in. Moreover, Seoulites really, seriously, desperately don't want to move out. The government has tried many times to create new cities away from Seoul with stable public-sector jobs; but a lot of people who work in Sejong, for example, just keep living in Seoul and make the 250km round trip every day. The 300K who actually moved to Sejong didn't make a dent in housing prices anywhere else, either.
pigsty 4y ago
> Just announcing to pull a whole industry further away from Seoul would immediately drive up housing prices of the destination and collapse the housing prices in Seoul. That’s a win for the average person in Seoul. Housing prices are destroying this generation of young people.

seoulmetro 4y ago

They have tried building hubs further out of Seoul, but it just doesn't work. Why would people leave and tie a billion liabilities to their workplace which is not even easy to maintain in Seoul itself? They want workers and they want Seoul. Workers want Seoul and they don't want to rely on their employers. This is always going to continue the fueling of Seoul border growth.

bobthepanda 4y ago
But workers want Seoul because already all the eggs are in the Seoul basket. Universities? Seoul. Jobs? Seoul. Services? Seoul. The country has been overconcentrated in Seoul since at least the beginning of the dictatorship. Breaking the cycle will not happen overnight, and you don’t even need to build new cities, just invest in the other ones. (Sejong was probably a bad idea.)
- seoulmetro 4y ago
  It's the same in every country. It's just that Korea started so late that there were no other cities to build from. Busan isn't a great city to grow.