True, but the people allocating compensation at TSMC say a few dozen steps away from the western money (which itself is at a disadvantage compared to an authoritarian regime when it comes to allocating large sums of money to arbitrary things) whereas the hiring manger for the other guys is only a couple layers of the org chart below someone who can personally direct the CCP money fire-hose. It's not like TSMC can just raise prices overnight to cover the expense. They'd have to get the money through some back-channel. Getting "state aid" from other states like that is something that is ridiculously tricky to pull off politically. TSMC is definitely at a disadvantage here.
OK, so 100k at most per engineer for "over 100" engineers, and those are just the ones that jumped ship, never mind the thousands staying behind. Is that significant?
Considering the average TSMC engineer makes 50k-60k, a 100k increase would probably be taken quite well.
Taiwan has such a big surplus of engineers and physicist that they don't really make much. Unless you are part of the club that knows TSMC special sauce in which you will be ridiculously well paid even by American standards, but that's not a large group.
> If central planning were a good way to run an economy, the Soviet Union would not have collapsed.
The Soviet Union put the first man into space and had severe shortages of food and consumer goods at the same time. Because that's what central planning does. It causes priorities to be distorted. Whatever the central planning committee is focused on happens and everything else withers and suffocates.
On net it's a massive catastrophic loss, but the places where resources are directed, get them in abundance. (This is also why it's extremely susceptible to corruption and favoritism.)
I'd say it was a data point. There is probably more concrete lessons to be learned before giving up on (admittedly complex and fraught) central planning
Especially given the fact that we have, in theory, the order of magnitude of computing power necessary to calculate an economy. But to implement it well would certainly be the most complex and titanic project in the history of computing, and even then it might fail for some unknown reason.
Worse yet, a failure where the wrong people accumulate wealth will just look like planned corruption to the masses. Sadly one reason why central planning often is accompanied by censorship
It has often occured to me that one of the rationale to start a colony in Mars is that wealth (at least on Earth) is mainly distributed among the wrong people, which is why so many short-sighted decisions have been made collectively by us who dominate the planet. And real innovations and progressive things are super rare. And they are often accompanied by (unnecessary) tragedies or sufferings that don't have to be that way. Theatres and facades are the norms.
But I now have come to the realisation that wrong is perhaps only in the eye of the beholder.
And there will always be censorship in a society as long as you look hard enough for it. Probably a feature not a bug though.
Maybe it just demonstrates this is the best we get collectively with neural circuitries like ours.
Central planning is 100% an engineering problem. Mises didn't call it the "Calculation Problem" for nothing. As long as economic complexity grows slower than processing power, there will be a point where central planning will be doable.
Of course, there is also the problem of incentives, but in the Soviet Union the problem of incentives was really only relevant as far as higher-ups reporting data, Soviet workers worked as hard and as well as any.
That is not to say that central planning directed by a centralized Leninist state is something that I think is desirable (or even something that would be theoretically desirable), but massive computing power is a quantum shift in the possibility of centralized and decentralized planning. The fact that it worked as well as it did in the USSR through pencil and paper is stunning.
Computers can’t determine economic priorities. In a free market demand does. In a planned economy politics does.
Computers are essentially useless for a planned economy. All they can do is provide quicker orders for factories to produce the wrong things in the wrong quantities.
When Gorbachev took over, they realized the government actually had no mathematical model of how the economy of USSR worked. Politicians had been basically been allocating resources blind for decades. They tried to shift to a more data driven approach toward the final years of the USSR, but it was way too late to have an effect.
Basically USSR collapsed for trying to match USA by overspending. Do you realized that is going on with China as well. Given their wastage is several magnitude higher than most western countries, go check what is happening to their reserve in the last 3 years. Also, seems like mother nature is compounding those wastage as well. Maybe we get to see Chernobyl 2 and Agfhan like what happened to USSR?
Perhaps that has something to do with having lost millions during WW1, revolution and WW2. Eastern Europe and China were hit the hardest during WW2 and didn't have the massive US economy (which was unharmed in both WW1 and WW2) to help the rebuild (marshal plan). The young people who died during WW2 would otherwise have contributed to the economy during the 50s, 60s and 70s. In fact, you can still see the damage of WW2 in the Russian demographics today.
Of course it didn't help that the soviet politicians constantly mismanaged the economy at the same time. Also interestingly, Mao was constantly screwing up the economy with his uneducated policies that he was hugely unpopular internally with the CCP. However Mao was hugely popular with the common people that the CCP had to just put up with and put out the fires he starts. (reminds me of Trump today...)
That's super false. They went from a backwards agricultural society with actual serfs to an industrialized world power in 3-4 decades despite a nazi invasion killing 20 million russians.
That sounds like decades of economic success to me.
Every first world country went from a backwards agricultural society to a far more advanced one during that century. Look up the dust bowl or even better the state of US agriculture during the early 20s.
Without central planning, the people’s of the Soviet republics would not just have been free, they would have grown wealthier
Russia has an economy the size of New York state. In a country of over 140 million people, that is no one's definition of success. Soviets didn't leave them much beyond a decaying nuclear arsenal and bunch of ugly buildings.
Introducing modern technology into an economy can do that. Look at all the East Asian nations such as Korea. On the whole, Russia's communism seems to have been a big drag on progress.
Lots of countries have developed. South Korea went from bombed-out wreck to one of the richest countries of the world. There's nothing impressive about Soviet development.
I can't believe that in 2020 there are still people defending a regime that murdered over 50 million people and ultimately collapsed due to an inherently flawed economic philosophy. And the argument is that the USSR was "too capitalist", and that's why it failed.
I get trying to move the Overton window, but this is too much. I'm hoping this is sarcasm on overdrive and I just missed it.
I'm talking about resource allocation, not central planning.
If Pooh Bear decides tomorrow that China will learn how to build pool floats then China will learn how to build pool floats. Everyone falls in line. If the leader of a western democracy wants to do something similar they have to convince people to want to do it and that takes time, policy carrots/sticks and propaganda/indoctrination.
The US military budget for 2020 is 720 billion dollars. That doesn't even include DHS. Most of it's going to bullshit, like Abrams tanks that the army doesn't even want.
Our problem isn't that we're too free and liberal, our problem is that we're too crooked and stagnant.
Granted...my cynical side is tempted to say, this is a backhanded way to get proper economic stimulus injected into the economy from fiscal conservatives...but graft is graft and in general is less efficient that a proper stimulus.
> a disadvantage compared to an authoritarian regime when it comes to allocating large sums of money to arbitrary things
Please. The US has never had an issue allocating arbitrarily large amounts of money to random things- either as military spending or moonshot projects. The Apollo program costed the equivalent of about 150 billion in today's money.
TSMC isn't a US company. The US can't just pump TSMC full of money without causing geopolitical problems and a overcoming internal political opposition (also true for US companies). When there's a perceived need (e.g. the Soviets beating us to space) and a huge majority approves or wants it the US can spend tons of money but this issue isn't to that point.
The US just forced TSMC to stop doing business with Huawei, even though Huawei was TSMC's 3rd biggest customer or so. Didn't seem like US got a lot of pushback from that.
CCP state-sponsored hiring of top talent from TSMC is plenty of pushback.
An ASEAN-EU-US alliance to directly source cutting-edge research from TSMC like the defense industry sourced from Silicon Valley in the early days when it was just a bunch of scattered suburbs among fruit orchards for example, would pump sufficient money to increase retention via pay and benefits. Lots of problems with that approach, of course. But options exist in Cold War II.
at high grow stage new competitor can offer at least 2x salary and 10x stocks (also convince them that stock will grow more). for that reason they get billions of investing which they can truly spend for anything. how can you compete with such offers?
Think about millionaires who joined Snap, Uber, Lyft, Pinterest at the right time like this
But there is real demand for cutting edge semiconductors in China... Do you think people will be satisfied with 14nm SMIC chips forever? Paying $140K salary is peanuts compared to what experienced software engineers at Facebook/Google get, their junior positions pays that much. Or hell $400K salaries at some financial firms.
Their semiconductor industry will not be solely made up of TSMC engineers, the vast majority are going to be Chinese juniors who will learn from them and eventually replace them.
The market reality is that the source of the chips can vanish at any time due to political reason, that's not a market problem anymore, it's now a political problem.
Not feasible. You’d have to offer ‘everything’ to every single one of your employees to prevent them being head hunted, but the opposition only has to make a better offer to a few of them to get the inside track info they need. That massive asymmetry makes it almost impossible to stop the brain drain.
Underpaying your critical employees is not a good strategy long term, even though it might look good on the balance sheet short term.
Let the free market decide. At some point, other, less tangible factors would start playing, such as individual freedoms, quality of the environment, family friednliness, etc.
It’s not a matter of under paying them. There isn’t a market for thousands of TSMC engineers at 2x or 3x salary in China so that really isn’t the market rate. There’s only a market for 100 of them, but it mostly doesn’t matter so much which 100.
TSMC wouldn’t have to just offer say 4x salary to 100 of their engineers to prevent the brain drain, they have to give it to several thousand engineers, and then a China would just up their offer to 5x because their ability to pick off a subset of people gives them 10x or even 100x leverage over TSMC. There’s just no way for TSMC to compete on salary like that. The most they could do is make it costly enough that the Chinese only hire away fewer engineers, at crippling cost to themselves, but they can never cut it down to zero.
There are other factors for these engineers to consider. This outflow may be temporary: once the Chinese side decides that their immediate goals have been met, they either force the Taiwanese engineers to leave, or accept Chinese citizenship, or some other unpleasant thing.
What I am saying is that the "free market" is not always limited to money. There are other things that play the role. TSMC must find ways to keep the essential engineers, or they are going to lose. This applies to anyone: you have to adjust to the changing environment, or die. Find something that will keep the best ones that other companies can't offer. Or die, like the rest of the manufacturing industry (look, American manufacturing is too expensive, let's make it in China! look, American programmers are too expensive, let's open a dev center in Bangalore!).
On the other hand, I think this is a welcome outcome - the knowledge spreads out and levels the playing field. From a very simplistic and idealistic view, it is better to have multiple industrial and knowledge centers than have everything concentrated in one place or country.
Semiconductor technology are like the new nuclear weapons, of the upmost importance to national security. Not having access to this technology will greatly affect the economy of a nation. The sooner everyone plays nice, the better it is for everybody. The Huawei ban is dumb and actually hurts Taiwan's leverage over China in the long term.
I would probably disagree on nuclear weapons: the lack of access to the latest semiconductors cannot wipe the humanity out, it merely retards the development of indigenous industries (other than semiconductor).
Huawei ban is a message to the Chinese government about explicit spying - and will probably get reversed pretty soon. You can spy, but don't make it so obvious.
Interestingly, I think these bans might have another, unintended, consequence: they might trigger acceleration of locally developed alternatives. We might see more of the Chinese-developed processors and IP in general in the future.
But I do agree that the sooner everyone plays nice, the better it is for everybody. The problem is that there are plenty of forces that benefit from chaos and suffering.
Sometimes your opponent can outspend you at every step and is possibly more motivated to achieve its goals. For China this is a far more pressing matter than profit, you might even see it as desperation. So this is like wrestling a bone from a hungry dog's mouth.
On the other hand TSMC is a publicly traded company. There's a limit to what expenses they can justify in front of the shareholders. China can lure engineers away with 3x the salary probably because that was the limit TSMC settled on, not because that's the limit of what China is willing to pay.
I agree with you and I don't understand how salaries are exempt from the normal capitalism dogma about supply and demand.
If I go into a supermarket and the bananas are too expensive then I can either pay or not eat a banana. If salaries are increasing companies can either pay market rate or no.
(They can also train internally but that doesn't seem fashionable nowadays)
> salaries are exempt from the normal capitalism dogma
A nation state devoting their trillions to outbid companies is the exact opposite of capitalist dogma. (Even less capitalist than a monopolist conglomerate abusing their dominant market position to outbid/undercut you.)
Should the market also be protected from other unfair things, like venture-capitalist backed startups that operate at a loss poaching key employees from established firms?
It might be rare, but right now those engineers are super valuable, and are getting paid for it.
demand and supply.
you see the same thing happen with self driving engineers commanding million dollar packages because they were in such high demand and supply was so low.
What we're discussing is that it's hard for a company to pay the market rate when the competition is the CCP itself. The CCP would pay the equivalent of $20 million if it needed to - and what then?
I get the point that everyone is making, including the point that you're trying to make. But if I just take your question at face value, isn't the answer simple: from billions in profit they make per quarter.
So you want to drain your profit to match the salary? I would think it would cause issue with the shareholder and when you want to pay for other future non salary expense, but fine lets ignore that. But what if China simply offer more salary again ? Remember, china has way more money than TSMC or even the whole Taiwan.
Salary isn’t everything. I would pick a nicer place to work over one with the highest salary. Salary just has to be good enough so you don’t feel cheated or don’t get the opportunities you want in life.
this is a nation state vs a private albeit important company. 'compete' isn't a word i'd use here no matter the multiplier as it implies the game is fair.
It doesn't matter. There will be an oversupply of capacity and talents down the road. When the inevitable downturn comes you will see even more outflow. You are basically trying to fight entropy here.
The entire fricking world except Intel gets their cutting-edge (read: profitable) chips made by TSMC. There is quite frankly a lot of latitude for them to raise prices.
Specializing in hardware design is really a poison pill despite how interesting the field is. You spend more than a decade in school to gain the skills needed to work at the chip level, which often requires a PhD. Then it turns out that there's only a handful of companies in the world with the equipment or supporting personnel needed for the field, and your skills are worthless outside of these specialized environments. No wonder pay is so low despite the years of schooling needed. After going through those considerations, I ended up dropping out of a PhD in RFIC design and fell back to embedded systems. More companies hiring and versatile enough to slide into general software or web programming if necessary. Chip design is still fascinating, but all traps look nice on the outside.
Also, for any aspiring undergrads wandering this way that thought RFIC design was fascinating because they picked up a copy of Razavi's "RF Microelectronics" or something similar and couldn't help but read it cover from cover, the textbooks lie. The analytical equations in a chip design textbook are not reflective of what the actual process is like. While you're still in college, try and get a professor to give you a student addition of Keysight ADS or Microwave Office AWR, along with an SDK from a fab. Then spend a few months of simulation time to see for yourself how much simulated results diverge from analytical models and how painstakingly slow it is to optimize those simulations especially once you start moving into EM simulations. And keep in mind these are simulations, chances are the physical chip will diverge even more due to process variations, unknown parasitics, or even bad simulation models from the fab. If the lack of precision, simulation time, and general drudgery doesn't kill your enthusiasm when compared to your textbook learning, then by all means proceed.
It's still a fascinating subject. Every so often when I need a break from code I'll take out my old RFIC textbooks and read a chapter or two. But you really don't want to know how this sausage is actually made.
I spent my teens looking up to microelectronics career, and actually self-studied a year, or two ahead for entrance to Nanyang, or NTU semi.eng. programs.
I almost accidentally met two ex-TSMC engineers who worked for Chang almost since his first days at ITRI, and they were very good mentors, and seriously warned me about "a decade+ grind ahead." In the end it were my parents who stonewalled the idea, and sent me to a business degree mill in Canada. And this is all was when I had 2 early admission letters from both Nanyang, and NTU on my desk...
> If the lack of precision, simulation time, and general drudgery doesn't kill your enthusiasm when compared to your textbook learning, then by all means proceed.
Semiconductor foundry is the current bottle neck of China. The agreement with U.S is lethal for mainland chip manufactures and will potentially grind the electronics market down to a halt. 4 of top 10 pure-play companies in the world is in Taiwan. IMO this move too obvious and weak to attract talents from these companies. Few people would join the 'chase up to the frontier of technical know-how' just because of better salary packages. It's possibly a 'worst case' scenario type of planning.