Fabs don’t require software engineers. They require materials and mechanical/electrical engineers, but mostly skilled technicians. Totally different education and skill set.
Also... 300-500K isn’t a mid-level programmers salary (you have been in San Francisco to long).
It's funny, because I had a similiar pitch back when everything was going down with Hong Kong to try and convince as many people and financial institutions to flee to Vancouver/Calgary.
Honestly the Federal government in Canada has never seemed willing to really actively take advantage of high immigration rates, to deliberately either steal firms or strip authoritarian countries of talent. Instead they end up passively being the consumer of last resort for PhD students and engineering grads from US universities who can't get green cards. Which is good, but feels like a missed opportunity.
My background is working in, and managing large wafer fabs in the US. The issue of staffing a fab is not well understood outside of the industry. There are three different workforces to assemble: semiconductor manufacturing engineers, maintenance technicians and operators. I started as a manufacturing engineer. I have a PhD in physics.
I had to acquire an eclectic group of EEs, MEs, materials engineers, chemical engineers, chemists and physicists as manufacturing engineers. Advanced degrees preferred. I had to have a sufficiently wide bandwidth of technical knowledge/skills to apply to an insanely complex assemblage of machines, instruments and processes. Infuriatingly, semiconductor processes tend to be meta stable. That is, they only work repeatably and consistently when the process variables are maintained precisely. The trick is identifying and understanding the process variables and how to control them. Statistical process control grew up in a fab. Our motto: Mother Nature hates semiconductor processing. One final comment. No fresh college grad arrives ready to work. Their real education begins when they hit the fab floor.
The complexity of the fab tools is staggering: high voltage plus high vacuum (ion implanters); extreme sensitivity to vibrations and particulate contamination (lithography); high temperature and precise temperature control (diffusion furnaces); exotic chemistry management (chemical vapor deposition and metal stutterers). The need for highly skilled and knowledgeable technicians is is very difficult to satisfy and maintain. Especially for 24/7 operations. The fab itself requires support for air management and utilities. The air at the fab floor must reliably not exceed having one particle larger than one micron in size per cubic meter!
Finally, the operators. They are the fuel of the fab. All of the engineers and technicians are in the fab to support their work. They need to be smart, detail oriented and diligent. Often they are the key to detecting minor deviations that can lead to big problems if not attended. Not surprisingly, the manufacturing engineers cultivate alliances with the operators. These informal teams are essential to smooth operations. Operators don’t have to have technical backgrounds, but when persons have two year or bachelor degrees, they are especially prized.
Fabs can’t work well in the boondocks. They require vast amounts of power, water and specialty gases. The equipment needs regular attention by the suppliers, so proximity to a large airport is very important.
I've heard from YIMBYs up there that things aren't _great_. Vancouver and Toronto are both pretty hard to build in, and same for the nearby suburbs. But I bet you could do something on open land near Winnipeg pretty easily.
The downside of course would be: Then you would be in Winnipeg. (It's beautiful, but there's a reason people also jokingly call it "Winterpeg".)
Eastern Alberta does not have plenty of water. You probably meant "Ontario and Manitoba."
Alberta is not in Central Canada. Manitoba really isn't either, by most definitions of the term.
Yes, in terms of immigration, Canada is more reminiscent of the bustling US before the 1930s.. Those immigrants almost exclusively go to Toronto, Montreal and Vancouver. None of them are very seismically stable, although Toronto is the best on that front. Some of Greater Toronto's satellite cities however are in Canada's tornado alley.
The Canadian shield is quite seismically stable. It's not close to Vancouver or Toronto.
Two important factors not considered in the above:
- Move to Montreal, and any immigrants would need to speak French in the workplace, because it has more than 25 employees.
- If you want to build a project of this size, it will attract the interest of the First Nations industry. If you think it takes a long time to build in the USA, in Canada First Nations have to be consulted, especially anywhere on or near the Canadian shield, and they have a de facto veto. Getting anything done would be 10-15 years in court. For large-scale projects to take place in Canada, governments must negotiate with First Nations, and both are incentivized to "negotiate", but never settle. So large-scale private projects are almost impossible.
To give you an idea of how bad it is, two examples are:
- in 2015, neither Canada nor the US had LNG terminals. With North America's shale revolution generating copious excess production, the scramble was on. Since then, the US has built 6 terminals, much of which will now go to Germany. Canada might get one terminal finished by 2025; the incomplete pipeline is subject to blockades and industrial sabotage.
- Canada is home to most of the metals needed for the EV revolution at global scale, and in deposits so profitable, they could vault Canada's GDP by up to 10% for decades. Developers have been trying to get mines operating since 1998, but have faced blockades, kidnappings, murders, FN de facto vetoes and a plethora of court cases for 25 years. More than one operator has lost >$500M in development costs before giving up on governments and FNs ever getting serious.
The US is nowhere near as development friendly as Taiwan or South Korea. Notice how well both have developed compared to North America since 1960. But for decades, Canada has been far, far less development friendly than the US.
Canada is disappointingly a build-nothing country as well. It may be politically a little different mix, but the overall vibe is a center-left government that pays lip service to key issues, but equally incapable of actually organizing people to make progress. Vetocracy all the way. It's a great idea were the existing residents interested in organizing and building a new industry. From what I see here on the ground, the immigration policy is not connected to any sort of industrial policy.
I don't see why this wouldn't work. But as a Canadian I'm not a fan of the idea of Canada as America-Tech-lite where workers can be paid a lot less (in high COL cities).
The Canadian government is glad to maintain these ludicrous immigration levels since the gov is captured by large business interests who want wages suppressed (and also more consumers for our oligopolies in telecom, groceries, banking, etc). Many Canadians, especially younger ones who are less likely to have secure housing, don't support that level of immigration.
The trope of Canadian tech workers being cheaper to pay is insulting, especially in the context of how high the cost of living is in Vancouver and Toronto. Many Canadian tech workers take remote US jobs, or just move to the US.
US, Canada plan North American chip corridor, starting with IBM expansion
March 24 (Reuters) - The United States and Canada said on Friday they would work together to create a bilateral semiconductor manufacturing corridor, as International Business Machines (IBM.N) signaled its intent to expand in Canada.
[...]
The Canadian government will spend C$250 million ($181.94 million) on its domestic semiconductor industry to boost research and development and manufacturing, the prime minister's office said in a statement.
For Canada, Waterloo is probably the best locale for this.
[...]
[According to CBRE, a larger share of Waterloo’s employment is high-tech than any of these cities. We didn’t cherry-pick, either. These comparators are among the top-ranked talent markets in the United States, according to . Waterloo has a greater density of tech talent than the Austin, Detroit or Columbus. We didn’t include it on the chart, but Waterloo also has a higher concentration of tech workers than Seattle or Denver, too. When it comes to talent density, Waterloo is the top mid-sized tech hub in North America.
[...]
One thing that really defines the is its sheer size. Yes, it’s highly rated, but it also has 10,000+ engineering students. It has about 4,000 computer science students. Among the comparison communities we listed above, it’s a clear leader, with only University of Michigan – Ann Arbor and Arizona State University – which has a very low global ranking for engineering – above us. No one comes close to UWaterloo’s computer science enrollment. It isn’t just about quantity, either (more about that on our next chart!). These are some of the best tech students you’ll find in North America and just about every graduate finishes with That helps explain why it’s one of Silicon Valley’s top recruitment schools.]
Waterloo is an hour's drive from Toronto, and looking to attract industry exactly as this blog proposes. The area is already deeply into photovoltaic R and D:
Centre for Advanced Photovoltaic and Display Systems (CAPDS) at the University of Waterloo.
[CAPDS promotes cutting edge research, training, and technology development in photovoltaic energy conversion and advanced display systems. We aim to be part of the solutions for these key sectors of the future.
Located at the University of Waterloo (UW), right at the heart of Canada's Technology Triangle area, the CAPDS is a 14000 square-foot state-of-the-art research facility with dedicated infrastructure for materials and devices research in the fields of photovoltaics and displays.]
Overall i suspect a chip fan would look a lot more like am automotive assembly plant from a workforce structure than a software company. Southwestern ON would be good from a land, water, low but not quite shield low seismic activity.
Alberta is more water constrained but still possible (and cheaper).
I don't think anywhere else has the pop base and colleges / universities to really train out a workforce.
It terms of building something, Ontario would be all over something equivalent to a new auto plant, esp in a different industry if it were a serious possibility. I doubt it has been considered serious so far though.
Excellent analysis of how 🤔 the chip problems could be a win-win for everyone 😀 if they would just common sense instead of greed determining how to have our own stable base of chip manufacturing here in North America 🌎 🇺🇸
Fabs don’t require software engineers. They require materials and mechanical/electrical engineers, but mostly skilled technicians. Totally different education and skill set.
Also... 300-500K isn’t a mid-level programmers salary (you have been in San Francisco to long).
But aside from that, it’s a good idea.
It's funny, because I had a similiar pitch back when everything was going down with Hong Kong to try and convince as many people and financial institutions to flee to Vancouver/Calgary.
Honestly the Federal government in Canada has never seemed willing to really actively take advantage of high immigration rates, to deliberately either steal firms or strip authoritarian countries of talent. Instead they end up passively being the consumer of last resort for PhD students and engineering grads from US universities who can't get green cards. Which is good, but feels like a missed opportunity.
My background is working in, and managing large wafer fabs in the US. The issue of staffing a fab is not well understood outside of the industry. There are three different workforces to assemble: semiconductor manufacturing engineers, maintenance technicians and operators. I started as a manufacturing engineer. I have a PhD in physics.
I had to acquire an eclectic group of EEs, MEs, materials engineers, chemical engineers, chemists and physicists as manufacturing engineers. Advanced degrees preferred. I had to have a sufficiently wide bandwidth of technical knowledge/skills to apply to an insanely complex assemblage of machines, instruments and processes. Infuriatingly, semiconductor processes tend to be meta stable. That is, they only work repeatably and consistently when the process variables are maintained precisely. The trick is identifying and understanding the process variables and how to control them. Statistical process control grew up in a fab. Our motto: Mother Nature hates semiconductor processing. One final comment. No fresh college grad arrives ready to work. Their real education begins when they hit the fab floor.
The complexity of the fab tools is staggering: high voltage plus high vacuum (ion implanters); extreme sensitivity to vibrations and particulate contamination (lithography); high temperature and precise temperature control (diffusion furnaces); exotic chemistry management (chemical vapor deposition and metal stutterers). The need for highly skilled and knowledgeable technicians is is very difficult to satisfy and maintain. Especially for 24/7 operations. The fab itself requires support for air management and utilities. The air at the fab floor must reliably not exceed having one particle larger than one micron in size per cubic meter!
Finally, the operators. They are the fuel of the fab. All of the engineers and technicians are in the fab to support their work. They need to be smart, detail oriented and diligent. Often they are the key to detecting minor deviations that can lead to big problems if not attended. Not surprisingly, the manufacturing engineers cultivate alliances with the operators. These informal teams are essential to smooth operations. Operators don’t have to have technical backgrounds, but when persons have two year or bachelor degrees, they are especially prized.
Fabs can’t work well in the boondocks. They require vast amounts of power, water and specialty gases. The equipment needs regular attention by the suppliers, so proximity to a large airport is very important.
Sorry to drone on. I think you get my point.
I've heard from YIMBYs up there that things aren't _great_. Vancouver and Toronto are both pretty hard to build in, and same for the nearby suburbs. But I bet you could do something on open land near Winnipeg pretty easily.
The downside of course would be: Then you would be in Winnipeg. (It's beautiful, but there's a reason people also jokingly call it "Winterpeg".)
A few geographical corrections:
Eastern Alberta does not have plenty of water. You probably meant "Ontario and Manitoba."
Alberta is not in Central Canada. Manitoba really isn't either, by most definitions of the term.
Yes, in terms of immigration, Canada is more reminiscent of the bustling US before the 1930s.. Those immigrants almost exclusively go to Toronto, Montreal and Vancouver. None of them are very seismically stable, although Toronto is the best on that front. Some of Greater Toronto's satellite cities however are in Canada's tornado alley.
The Canadian shield is quite seismically stable. It's not close to Vancouver or Toronto.
Two important factors not considered in the above:
- Move to Montreal, and any immigrants would need to speak French in the workplace, because it has more than 25 employees.
- If you want to build a project of this size, it will attract the interest of the First Nations industry. If you think it takes a long time to build in the USA, in Canada First Nations have to be consulted, especially anywhere on or near the Canadian shield, and they have a de facto veto. Getting anything done would be 10-15 years in court. For large-scale projects to take place in Canada, governments must negotiate with First Nations, and both are incentivized to "negotiate", but never settle. So large-scale private projects are almost impossible.
To give you an idea of how bad it is, two examples are:
- in 2015, neither Canada nor the US had LNG terminals. With North America's shale revolution generating copious excess production, the scramble was on. Since then, the US has built 6 terminals, much of which will now go to Germany. Canada might get one terminal finished by 2025; the incomplete pipeline is subject to blockades and industrial sabotage.
- Canada is home to most of the metals needed for the EV revolution at global scale, and in deposits so profitable, they could vault Canada's GDP by up to 10% for decades. Developers have been trying to get mines operating since 1998, but have faced blockades, kidnappings, murders, FN de facto vetoes and a plethora of court cases for 25 years. More than one operator has lost >$500M in development costs before giving up on governments and FNs ever getting serious.
The US is nowhere near as development friendly as Taiwan or South Korea. Notice how well both have developed compared to North America since 1960. But for decades, Canada has been far, far less development friendly than the US.
Canada is disappointingly a build-nothing country as well. It may be politically a little different mix, but the overall vibe is a center-left government that pays lip service to key issues, but equally incapable of actually organizing people to make progress. Vetocracy all the way. It's a great idea were the existing residents interested in organizing and building a new industry. From what I see here on the ground, the immigration policy is not connected to any sort of industrial policy.
I don't see why this wouldn't work. But as a Canadian I'm not a fan of the idea of Canada as America-Tech-lite where workers can be paid a lot less (in high COL cities).
The Canadian government is glad to maintain these ludicrous immigration levels since the gov is captured by large business interests who want wages suppressed (and also more consumers for our oligopolies in telecom, groceries, banking, etc). Many Canadians, especially younger ones who are less likely to have secure housing, don't support that level of immigration.
The trope of Canadian tech workers being cheaper to pay is insulting, especially in the context of how high the cost of living is in Vancouver and Toronto. Many Canadian tech workers take remote US jobs, or just move to the US.
Hahahaha you obviously don’t know much about Canadian red tape
Canada’s housing crisis is way, way worse than the US.
US, Canada plan North American chip corridor, starting with IBM expansion
March 24 (Reuters) - The United States and Canada said on Friday they would work together to create a bilateral semiconductor manufacturing corridor, as International Business Machines (IBM.N) signaled its intent to expand in Canada.
[...]
The Canadian government will spend C$250 million ($181.94 million) on its domestic semiconductor industry to boost research and development and manufacturing, the prime minister's office said in a statement.
[...]
https://www.reuters.com/technology/us-canada-plan-north-american-chip-corridor-starting-with-ibm-expansion-2023-03-24/
For Canada, Waterloo is probably the best locale for this.
[...]
[According to CBRE, a larger share of Waterloo’s employment is high-tech than any of these cities. We didn’t cherry-pick, either. These comparators are among the top-ranked talent markets in the United States, according to . Waterloo has a greater density of tech talent than the Austin, Detroit or Columbus. We didn’t include it on the chart, but Waterloo also has a higher concentration of tech workers than Seattle or Denver, too. When it comes to talent density, Waterloo is the top mid-sized tech hub in North America.
[...]
One thing that really defines the is its sheer size. Yes, it’s highly rated, but it also has 10,000+ engineering students. It has about 4,000 computer science students. Among the comparison communities we listed above, it’s a clear leader, with only University of Michigan – Ann Arbor and Arizona State University – which has a very low global ranking for engineering – above us. No one comes close to UWaterloo’s computer science enrollment. It isn’t just about quantity, either (more about that on our next chart!). These are some of the best tech students you’ll find in North America and just about every graduate finishes with That helps explain why it’s one of Silicon Valley’s top recruitment schools.]
[...]
https://www.waterlooedc.ca/blog/five-charts-waterloo-tech-hub
Waterloo is an hour's drive from Toronto, and looking to attract industry exactly as this blog proposes. The area is already deeply into photovoltaic R and D:
Centre for Advanced Photovoltaic and Display Systems (CAPDS) at the University of Waterloo.
[CAPDS promotes cutting edge research, training, and technology development in photovoltaic energy conversion and advanced display systems. We aim to be part of the solutions for these key sectors of the future.
Located at the University of Waterloo (UW), right at the heart of Canada's Technology Triangle area, the CAPDS is a 14000 square-foot state-of-the-art research facility with dedicated infrastructure for materials and devices research in the fields of photovoltaics and displays.]
[...]
https://www.capds.uwaterloo.ca/services.htm
Overall i suspect a chip fan would look a lot more like am automotive assembly plant from a workforce structure than a software company. Southwestern ON would be good from a land, water, low but not quite shield low seismic activity.
Alberta is more water constrained but still possible (and cheaper).
I don't think anywhere else has the pop base and colleges / universities to really train out a workforce.
It terms of building something, Ontario would be all over something equivalent to a new auto plant, esp in a different industry if it were a serious possibility. I doubt it has been considered serious so far though.
USA: we will mandate that new semiconductor factories must provide childcare facilities.
Canada: everywhere has $10/day childcare already.
The USA is so afraid of universal benefits that it will twist itself in knots to get more places with benefits....
Excellent analysis of how 🤔 the chip problems could be a win-win for everyone 😀 if they would just common sense instead of greed determining how to have our own stable base of chip manufacturing here in North America 🌎 🇺🇸
Ontario (especially Toronto) is just as hellish to build anything as it is in the US. Maybe in the eastern provinces this might work.
With climate change, Arizona summers are going to be nasty. And there are serious questions about water supply.