Can Semis Save the Economy?

[Arthur Hanson]

Platforms, physical and virtual, being integrated into economic structures offer not only the US, but the world a new way forward if entrenched special interests up and down our economic chain can be overcome. This will require a major change in which we fully utilize the still untapped power of semis' vast power to transfer, store and process literally everything, leveraging time and resources on a scale that even now is unimaginable to most. Just in diagnostics and monitoring of everything from education, training, manufacturing, medical, and almost every human endeavor the potential of platforms and procedures enabled by advanced semi and sensors ever-increasing capabilities at ever lower costs promise an economic and social revolution in much more effective and productive use of our resources and time. Special interests at every level and field are putting up roadblocks to protect systems that have become outdated to the extreme. Small individual practices in many fields and endeavors have models that are grossly inefficient and ineffective in the application of time and resources to the extreme. Semis of all types offer us the opportunity to achieve vastly improved use of time, talent, and resources while delivering superior services and products. Semis and the ecosystems they have created now allow us to leverage resources of all types to multiples that at one time were undreamed of and now possible also to apply in ways every day as their power, memory, versatility at ever lower costs put staggering opportunities in front of us, just waiting for our actions. We need people of vision and courage to step forward and realize we literally need a revolution to overcome the special interests that have given us the current mess we have that is not only at our expense but their own if they are smart enough to realize it. I just got done listening to CEOs talking on STEM education and they found that STEM students have double the earning power of non-STEM students, which by definition means they are making a much larger contribution to moving society forward. Removing commercial and physical conflict is the starting point and to realizing people that promote it are a threat to us all and the world, which is getting smaller by the day. Education, medical and the military are the places to start for their massive consumption of time and resources. Any thoughts or comments appreciated.

 

[Daniel Nenni]

Yes. Semiconductors have been driving the economy since the smartphone.

 

[tooLongInEDA]

Platforms, physical and virtual, being integrated into economic structures offer not only the US, but the world a new way forward if entrenched special interests up and down our economic chain can be overcome. This will require a major change in which we fully utilize the still untapped power of semis' vast power to transfer, store and process literally everything, leveraging time and resources on a scale that even now is unimaginable to most. Just in diagnostics and monitoring of everything from education, training, manufacturing, medical, and almost every human endeavor the potential of platforms and procedures enabled by advanced semi and sensors ever-increasing capabilities at ever lower costs promise an economic and social revolution in much more effective and productive use of our resources and time. Special interests at every level and field are putting up roadblocks to protect systems that have become outdated to the extreme. Small individual practices in many fields and endeavors have models that are grossly inefficient and ineffective in the application of time and resources to the extreme. Semis of all types offer us the opportunity to achieve vastly improved use of time, talent, and resources while delivering superior services and products. Semis and the ecosystems they have created now allow us to leverage resources of all types to multiples that at one time were undreamed of and now possible also to apply in ways every day as their power, memory, versatility at ever lower costs put staggering opportunities in front of us, just waiting for our actions. We need people of vision and courage to step forward and realize we literally need a revolution to overcome the special interests that have given us the current mess we have that is not only at our expense but their own if they are smart enough to realize it. I just got done listening to CEOs talking on STEM education and they found that STEM students have double the earning power of non-STEM students, which by definition means they are making a much larger contribution to moving society forward. Removing commercial and physical conflict is the starting point and to realizing people that promote it are a threat to us all and the world, which is getting smaller by the day. Education, medical and the military are the places to start for their massive consumption of time and resources. Any thoughts or comments appreciated.

I'd love to agree that STEM students make a large contribution to society, but your suggestion that there is some linear relationship between what people are paid (note: paid, not earn) and their contribution to society seems flawed. There are countless counter-examples and many vital but low paid jobs where the supply/demand balance (amongst other factors) suppresses wages.

But you are correct that the role of semis is to enable cost and efficiency savings in other industries and activities. And that not all those fields are currently taking advantage of the opportunities (and hence not passing on the benefits to their customers and the taxpayers who frequently support them).

Having said all that, I am not convinced that the current generation of "disruptive" start-ups are always the ones doing the useful and necessary work here.

Here's one example of activities where practice seems to be stuck in the dark ages (and this is nothing to do with semis). In both parliament (in your case the Senate and Congress), debate is still strictly verbal. You might imagine that visual aids and multimedia could be used to explain complex arguments and data here. But no. They are still in the eighteenth century. The same seems to be true of law courts. Couldn't they use logical flowcharts to help jurors trace from evidence, assumptions and inferences to conclusions ?

 

[hist78]

I'd love to agree that STEM students make a large contribution to society, but your suggestion that there is some linear relationship between what people are paid (note: paid, not earn) and their contribution to society seems flawed. There are countless counter-examples and many vital but low paid jobs where the supply/demand balance (amongst other factors) suppresses wages.

But you are correct that the role of semis is to enable cost and efficiency savings in other industries and activities. And that not all those fields are currently taking advantage of the opportunities (and hence not passing on the benefits to their customers and the taxpayers who frequently support them).

Having said all that, I am not convinced that the current generation of "disruptive" start-ups are always the ones doing the useful and necessary work here.

Here's one example of activities where practice seems to be stuck in the dark ages (and this is nothing to do with semis). In both parliament (in your case the Senate and Congress), debate is still strictly verbal. You might imagine that visual aids and multimedia could be used to explain complex arguments and data here. But no. They are still in the eighteenth century. The same seems to be true of law courts. Couldn't they use logical flowcharts to help jurors trace from evidence, assumptions and inferences to conclusions ?

There are 435 members in the Congress. If each one of them gives an one-minute speech during one single legislation debate, it will take at least 10 hours to finish (assume there is a 30 seconds walking/transition time in between each speaker).

Yes, we need each representative to cover a relatively smaller congressional district than the senators. But to expect this structure to function in the 21st century is really questionable.

 

[tooLongInEDA]

There are 435 members in the Congress. If each one of them gives an one-minute speech during one single legislation debate, it will take at least 10 hours to finish (assume there is a 30 seconds walking/transition time in between each speaker).

Yes, we need each representative to cover a relatively smaller congressional district than the senators. But to expect this structure to function in the 21st century is really questionable.

There are over 1000 in the UK House of Lords - and Boris Johnson is about to add a whole load more ! Plus 660 MPs in the Commons. Not bad for a small country ... but in reality probably 3x what it should be.

 

[Arthur Hanson]

I'd love to agree that STEM students make a large contribution to society, but your suggestion that there is some linear relationship between what people are paid (note: paid, not earn) and their contribution to society seems flawed. There are countless counter-examples and many vital but low paid jobs where the supply/demand balance (amongst other factors) suppresses wages.

But you are correct that the role of semis is to enable cost and efficiency savings in other industries and activities. And that not all those fields are currently taking advantage of the opportunities (and hence not passing on the benefits to their customers and the taxpayers who frequently support them).

Having said all that, I am not convinced that the current generation of "disruptive" start-ups are always the ones doing the useful and necessary work here.

Here's one example of activities where practice seems to be stuck in the dark ages (and this is nothing to do with semis). In both parliament (in your case the Senate and Congress), debate is still strictly verbal. You might imagine that visual aids and multimedia could be used to explain complex arguments and data here. But no. They are still in the eighteenth century. The same seems to be true of law courts. Couldn't they use logical flowcharts to help jurors trace from evidence, assumptions and inferences to conclusions ?

You are dead on, the court and legal system have turned into a nightmare of not only inefficiency and waste but in many cases injustice. In many cases, it's money that rules and wins, when the most important issue is the facts, not the best talking head lawyer. I fully agree with your suggestions for the legal system, they would deliver better and less biased results with a large savings in time and cost. Our development of technologies are definitely outpacing our ability to integrate and use them. Much of this is due to special interests that are only interests are their own, to the detriment of society at large.

 

[Paul2]

I'd love to agree that STEM students make a large contribution to society, but your suggestion that there is some linear relationship between what people are paid (note: paid, not earn) and their contribution to society seems flawed. There are countless counter-examples and many vital but low paid jobs where the supply/demand balance (amongst other factors) suppresses wages.

The Western narrative about manufacturing is wrong, totally wrong. More STEM != manufacturing returning to the West.

More engineers != more widgets. More workers is more widget, but not all 6 years degree holding engineers are willing to work on the assembly lines.

12 years ago, we had excellent assembly line workers, because for $400-$600 per month you were able to hire a degreed engineer with some experience to work on the line. The best of them were then promoted to line, and then shift, and then factory level managers.

What's happening now? There are no workers for these engineers to manage. A $3000 per month shift manager can still live in Shenzhen, but regular workers only survive, and they leave the city. Talented people don't get to work on the line anymore, but they go for extraordinary effort to get these few $3000-per-month manager jobs remaining. We literally had cases of HR manager being bribed, forged degrees, and references.

The best indicator for long term viability of the place for manufacturing industry is the cost of living, and life prospects for an educated labourer. Cheap housing it the best way the govt can subsidise manufacturing.

12 years go, line work was a dream job for an average youth in south China: you will be able to buy a motorcycle from your first salary, get a girlfriend at the factory, pay for a tech college in 1 year, live in a free dorm space being fed, and boarded by employer, get the company pay for your certification, or upskilling course, and move out of a town in 4-5 years as a big man, or, if you are skilled. and lucky, settle in the city.

And the city was very welcoming, it was the best city in the country to be a 20 years old: cheap night life, snacks, and food for a working man on every corner. Cheap illegal dorms were overlooked, and much of the garage industry too. Motorcycle ownership was not curtailed, or looked upon as it was in the rest of the country. The inter-provincial racism was also completely unheard of, which was unique for China (Chinese distaste immigrants from inner provinces more than foreign immigrants.)

Now, there is an aura of hopelessness over the industry. Most of workers are middle age men who never managed to make it anywhere else. They are 2-4 times less productive than young, bright people we had 12 years ago.

Reason for this all: Shenzhen allowed the cost of housing to past Manhattan, and overfocused on attracting the elite managers, and went extra hostile on blue collar lifestyle.

But you are correct that the role of semis is to enable cost and efficiency savings in other industries and activities. And that not all those fields are currently taking advantage of the opportunities (and hence not passing on the benefits to their customers and the taxpayers who frequently support them).

The role of subsidies is misunderstood by Western policymakers when it comes to tech. Very often subsidies only incentivise continued overreliance on inherently more expensive, inefficient, or overkill technologies, vs. allowing a superior, cheaper alternative come.

Germany erased itself from an IT industry map by subsidising copper based telecoms.

Sematech KILLED the remaining US semis largely because they weren't pushed to move to CMOS early enough. Instead, Sematechers wasted the money on salvaging their existing process, designs, and products, or just took the money, and did nothing at all.

China began chasing USA in the high status high tech niches and abandoned its strength in down to earth manufacturing approach. Chinese manufacturers pocketed all the "high-tech" subsidies, and did an out of tact hightechisation which crippled them. I would say, Chinese indigenous "Shanzhai" tech was already dead by around 2015 on the enormous wave of subsidy receivers crowding out real innovators.

Manufacturing people invariably use government handouts as a leverage for existing business, rather than internal investments. Cashing out subsidies pay, and the investment for future would not before you are crowded out by people who were able to make a quick buck.

Only by sensible deregulation, and agile adaptability can the government benefit the business. Businesses produce value — Governments take value, and the only way the can help is by finding out how to not to destroy what's produced. Western policy making elites don't understand the concept of delayed gratification.

The biggest, and single most harmful impediment for manufacturing in the West is its legal-regulatory system being turned into a minefield.

Having said all that, I am not convinced that the current generation of "disruptive" start-ups are always the ones doing the useful and necessary work here.

Agreed. The silicon valley is a hollow cult parasitising on the ultra low rates. Stock market sucks life out of the real market.

 

[jms_embedded]

Sematech KILLED the remaining US semis largely because they weren't pushed to move to CMOS early enough. Instead, Sematechers wasted the money on salvaging their existing process, designs, and products, or just took the money, and did nothing at all.

Interesting... where could I find more information?

 

[Paul2]

Interesting... where could I find more information?

Nothing written on it in wider literature, but I spoke with 2 semi engineers who came with Morris to ITRI, and then TSMC around a decade ago, and plainly asked why American semi industry went belly up.

Motorola was making tons of money selling respins of close to a decade old designs (defence contracts) on depletion mode logic, or even NPNs, and wanted to extend their life. When they finally admitted them being a dead-end, they only had 1st generation CMOS for new products (the need for which exploded with the advent of PCs) And Motorola was still very advanced in comparison to, say, Sinclair, RCA, or other 2nd tier semis which weren't run by engineers, and who had no idea what is coming.

 

[chris.sparks]

Nothing written on it in wider literature, but I spoke with 2 semi engineers who came with Morris to ITRI, and then TSMC around a decade ago, and plainly asked why American semi industry went belly up.

Motorola was making tons of money selling respins of close to a decade old designs (defence contracts) on depletion mode logic, or even NPNs, and wanted to extend their life. When they finally admitted them being a dead-end, they only had 1st generation CMOS for new products (the need for which exploded with the advent of PCs) And Motorola was still very advanced in comparison to, say, Sinclair, RCA, or other 2nd tier semis which weren't run by engineers, and who had no idea what is coming.

I too am curious about the SEMATECH comments. My recollection is that we were focused on precompetitive R&D, tool development, standardization, etc. at least in when they were in Austin. The point wasn't to develop CMOS (unless for test structures?). It's been about 15 years since I worked for them so my memory may be on the hazy side and I've not followed them much since they moved to Albany.

 

[Paul2]

I too am curious about the SEMATECH comments. My recollection is that we were focused on precompetitive R&D, tool development, standardization, etc. at least in when they were in Austin. The point wasn't to develop CMOS (unless for test structures?). It's been about 15 years since I worked for them so my memory may be on the hazy side and I've not followed them much since they moved to Albany.

One way or another, the overwhelming wave of government grants in eighties on the wave of Japan scare lead to crowding out of real innovators who weren't as lucky getting grants, and scoring buzzwords. It lead to overfocus on cash crop chips, where the immediate benefit was the highest, while Asia pioneered both the low profit VLSIs, and low-end ICs on relatively new nodes, which eventually became Taiwan's strength.

US semi industry was already on CMOS (and really ahead of Japan for the top tier) by eighties, and the last thing it needed was an incentive to hold back on transition, and to milk existing fab lines to the end.