If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Just hope China doesn't do anything silly with Taiwan.
My day job is as a semis / electronics industry analyst. I've never seen a shortage like this one in over 30 years because it is so multi-faceted.
On the supply side it's not just leading edge foundry capacity that is tight; 8 inch legacy capacity is also fully booked. And in packaging, assembly and test, as well as tight capacity, there are also shortages of raw materials (lead frames, PCBs for advanced packaging, etc). We've also had the Texas freeze shutting down Samsung, Infineon and NXP fabs for a month and a fire at Renesas in Japan also shutting it down for a few weeks.
On the demand side, in addition to covid-related demand from hyperscalers (cloud servers and storage) and work/study from home PCs/Chromebooks, we also have the rapid migration to 5G smartphones (>500m units this year) and the rebound in automotive (including the shift to EV). 5G and EV need LOTS more chips per system for functions like RF and Power Management.
So, this situation isn't getting better any time soon - despite increased CAPEX now, it's takes a couple of years to bring a new fab up to full production. Wafer and chip ASPs are up significantly in 1Q and lead-times are out to a year for some chips. Double and triple ordering is rife. This situation will continue through 2021 and get gradually eased in 2022. The semis market will grow 20% this year, much of the revenue growth fueled by chip price rises.
The US is seriously worried (belatedly) about the national security implications of all this - i.e. the reliance on TSMC in Taiwan to build the world's semiconductors and the ambitions of China to develop it's own or "acquire" the technology it needs to build a competing technology roadmap over the next decade or two. EUV is the jewel in the crown here - without that, China will struggle to go head to head with TSMC, Samsung/SK hynix and Intel.
Hence the money being showered on Intel and TSMC by Biden to build fabs in Arizona ... this is about national security ("techo-nationalism" if you like). The globalized supply chain is going to be re-engineered, at least to some degree.
The interesting play here is the EU ... which way does it jump? Does it align with the US camp or cosy up to China? ASML (the only EUV supplier) is of course a European company ...
I suspect the EU won't actually do anything useful - any attempt to build fabs in within the EU will lead to years of arguments over where they should be built and who is subsiding them.
I do wonder if the UK are paying any attention at all on working how to get some of those plants in a globally focussed country with little chance of natural disaster or regional wars.
Gelsinger has been taking to "the European Commission" about a new Intel fab in the EU (not Ireland) ... smart money would be on The Netherlands or East Europe.
You'd hope the UK would be looking at it because we have so many in-built advantages (political stability, no natural disasters, English-speaking, good Universities, good global trade links, etc, etc) but I suspect the focus is more on life-sciences and green tech.
Just hope China doesn't do anything silly with Taiwan.
My day job is as a semis / electronics industry analyst. I've never seen a shortage like this one in over 30 years because it is so multi-faceted.
On the supply side it's not just leading edge foundry capacity that is tight; 8 inch legacy capacity is also fully booked. And in packaging, assembly and test, as well as tight capacity, there are also shortages of raw materials (lead frames, PCBs for advanced packaging, etc). We've also had the Texas freeze shutting down Samsung, Infineon and NXP fabs for a month and a fire at Renesas in Japan also shutting it down for a few weeks.
On the demand side, in addition to covid-related demand from hyperscalers (cloud servers and storage) and work/study from home PCs/Chromebooks, we also have the rapid migration to 5G smartphones (>500m units this year) and the rebound in automotive (including the shift to EV). 5G and EV need LOTS more chips per system for functions like RF and Power Management.
So, this situation isn't getting better any time soon - despite increased CAPEX now, it's takes a couple of years to bring a new fab up to full production. Wafer and chip ASPs are up significantly in 1Q and lead-times are out to a year for some chips. Double and triple ordering is rife. This situation will continue through 2021 and get gradually eased in 2022. The semis market will grow 20% this year, much of the revenue growth fueled by chip price rises.
The US is seriously worried (belatedly) about the national security implications of all this - i.e. the reliance on TSMC in Taiwan to build the world's semiconductors and the ambitions of China to develop it's own or "acquire" the technology it needs to build a competing technology roadmap over the next decade or two. EUV is the jewel in the crown here - without that, China will struggle to go head to head with TSMC, Samsung/SK hynix and Intel.
Hence the money being showered on Intel and TSMC by Biden to build fabs in Arizona ... this is about national security ("techo-nationalism" if you like). The globalized supply chain is going to be re-engineered, at least to some degree.
The interesting play here is the EU ... which way does it jump? Does it align with the US camp or cosy up to China? ASML (the only EUV supplier) is of course a European company ...
I suspect the EU won't actually do anything useful - any attempt to build fabs in within the EU will lead to years of arguments over where they should be built and who is subsiding them.
I do wonder if the UK are paying any attention at all on working how to get some of those plants in a globally focussed country with little chance of natural disaster or regional wars.
Gelsinger has been taking to "the European Commission" about a new Intel fab in the EU (not Ireland) ... smart money would be on The Netherlands or East Europe.
You'd hope the UK would be looking at it because we have so many in-built advantages (political stability, no natural disasters, English-speaking, good Universities, good global trade links, etc, etc) but I suspect the focus is more on life-sciences and green tech.
(I must admit my reaction to @RochdalePioneers doomladen agonising yesterday was to wonder when the data was coming out!)
Its only a 27% drop. A big success really.
Yes because February 2021 is definitely comparable to February 2020. I mean it's not as if anything is different now at all.
Whether that data is accurate we'll have to wait and see, but considering that we're in full lockdown in February, to have more exports in Feb 2021 on that data than any of May, June, July or August 2020 seems really surprising.
The comments on the union earlier (Fridays tend to be full of meetings) - it cannot be sustained in its current form. For all the flag twattery being done by Tories to gee up English votes, it has the opposite effect in the other nations.
A federal union may work. But I don't see anyone other than the LibDems seriously arguing for it. So England will try to colonially repress the savages in the provinces, but ultimately that only drives harder the forces tearing the union apart.
If England wants supremacy to do what it wants without regard to what the people it trades with wants, then sadly we should let it.
I am now hoping that they don’t bat as abjectly as the did last time they faced Jamie Overton...
It's alright - we appear to not be bowling him first up so that your openers can get in...
They’ve not bowled too smartly so far, I’m afraid. Dent loves (a) width (b) balls on the pad. Best way to get him is to bowl tight just outside off, when he tries to cut and drags on. Plenty of the first two, not nearly enough to the last. He’ll leave balls as wide as the ones Topley bowled all day.
But it’s Glaws. They’ll find inventive ways to combust if given a little time.
Just hope China doesn't do anything silly with Taiwan.
My day job is as a semis / electronics industry analyst. I've never seen a shortage like this one in over 30 years because it is so multi-faceted.
On the supply side it's not just leading edge foundry capacity that is tight; 8 inch legacy capacity is also fully booked. And in packaging, assembly and test, as well as tight capacity, there are also shortages of raw materials (lead frames, PCBs for advanced packaging, etc). We've also had the Texas freeze shutting down Samsung, Infineon and NXP fabs for a month and a fire at Renesas in Japan also shutting it down for a few weeks.
On the demand side, in addition to covid-related demand from hyperscalers (cloud servers and storage) and work/study from home PCs/Chromebooks, we also have the rapid migration to 5G smartphones (>500m units this year) and the rebound in automotive (including the shift to EV). 5G and EV need LOTS more chips per system for functions like RF and Power Management.
So, this situation isn't getting better any time soon - despite increased CAPEX now, it's takes a couple of years to bring a new fab up to full production. Wafer and chip ASPs are up significantly in 1Q and lead-times are out to a year for some chips. Double and triple ordering is rife. This situation will continue through 2021 and get gradually eased in 2022. The semis market will grow 20% this year, much of the revenue growth fueled by chip price rises.
The US is seriously worried (belatedly) about the national security implications of all this - i.e. the reliance on TSMC in Taiwan to build the world's semiconductors and the ambitions of China to develop it's own or "acquire" the technology it needs to build a competing technology roadmap over the next decade or two. EUV is the jewel in the crown here - without that, China will struggle to go head to head with TSMC, Samsung/SK hynix and Intel.
Hence the money being showered on Intel and TSMC by Biden to build fabs in Arizona ... this is about national security ("techo-nationalism" if you like). The globalized supply chain is going to be re-engineered, at least to some degree.
The interesting play here is the EU ... which way does it jump? Does it align with the US camp or cosy up to China? ASML (the only EUV supplier) is of course a European company ...
I suspect the EU won't actually do anything useful - any attempt to build fabs in within the EU will lead to years of arguments over where they should be built and who is subsiding them.
I do wonder if the UK are paying any attention at all on working how to get some of those plants in a globally focussed country with little chance of natural disaster or regional wars.
Gelsinger has been taking to "the European Commission" about a new Intel fab in the EU (not Ireland) ... smart money would be on The Netherlands or East Europe.
You'd hope the UK would be looking at it because we have so many in-built advantages (political stability, no natural disasters, English-speaking, good Universities, good global trade links, etc, etc) but I suspect the focus is more on life-sciences and green tech.
Netherlands makes the most sense or simply expanding the Irish output but if it's definitely a new foundry it wouldn't make sense to do it anywhere other than the Netherlands. I doubt the UK would be in contention, we lack expertise in the sector.
I suppose when you've tried and failed with all the other radicalisations, right wing culture warrior is what's left. Galloway's novel is going to be an absolute stinker, pound shop orator soundbites repeated page after page.
(I must admit my reaction to @RochdalePioneers doomladen agonising yesterday was to wonder when the data was coming out!)
Its only a 27% drop. A big success really.
Yes because February 2021 is definitely comparable to February 2020. I mean it's not as if anything is different now at all.
Whether that data is accurate we'll have to wait and see, but considering that we're in full lockdown in February, to have more exports in Feb 2021 on that data than any of May, June, July or August 2020 seems really surprising.
Industry was properly shut down or heavily locked down in the first wave. Not so any more. This winter lockdown has been people focused rather than a general job stopped like the first time.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
(I must admit my reaction to @RochdalePioneers doomladen agonising yesterday was to wonder when the data was coming out!)
Its only a 27% drop. A big success really.
Yes because February 2021 is definitely comparable to February 2020. I mean it's not as if anything is different now at all.
Whether that data is accurate we'll have to wait and see, but considering that we're in full lockdown in February, to have more exports in Feb 2021 on that data than any of May, June, July or August 2020 seems really surprising.
Industry was properly shut down or heavily locked down in the first wave. Not so any more. This winter lockdown has been people focused rather than a general job stopped like the first time.
Yes, but May, June, July and August 2020 was less locked down here and abroad than Feb 2021.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
Would it make sense for the UK to invest in this unilaterally?
How much would it cost to become a leading chipmaker domestically? And given the national security implications if we were to contrast that with eg the cost of Trident how would it compare?
Having a domestic supply of chips seems like it could be a national security issue and an infinitely more useful thing economically than Trident.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
it wouldn't be as easy as S Korea or Taiwan found it a few decades back, either. Taiwan more or less invented the pure foundry model, which meant they got ever smart engineer form the US who wanted to start their own company turning up as a customer. We don't even have much a domestic market to support that, and nothing to attract overseas customers. Really difficult to see how we could do it without attracting major established manufacturers, I think ?
That said, it might still be a better use of money than Trident/Trident replacement (which might end up getting scrapped anyway should it lose its Scottish base).
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
I don't think either Philip or myself were suggesting we build an entirely new company, more that it may make sense to encourage TSMC, Intel or indeed both to build a fab or 2 in the UK.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
You'd be competing against plants where the capital expenditure was paid off, back when the process was the big thing, and the prices were high.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
I don't think either Philip or myself were suggesting we build an entirely new company, more that it may make sense to encourage TSMC, Intel or indeed both to build a fab or 2 in the UK.
Intel maybe, but TSMC is highly unlikely to build outside of Taiwan. We'd probably be better off incentivising Global Foundries to build a 10nm fabrication plant here, it would be something covered by our agreement with the EU for tariff/quota free exports, and benefit from our entry to the CPTPP as well.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
You'd be competing against plants where the capital expenditure was paid off, back when the process was the big thing, and the prices were high.
So I don't think it would ever turn a profit.
Yes, literally never. The £80bn would have to be written off.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
Not suggesting an entire industry from scratch, heck not even suggesting we actually do anything.
Just thinking out loud, the security implications of this seem concerning regarding China/Taiwan etc - plus then having what happened with vaccines etc over the past 12 months.
Obviously we don't want to on-shore everything, but having some form of something domestically made seems like it could be a better idea for economic and national security than eg Trident.
OTOH it could be a mammoth blackhole and a terrible idea.
I have no opinion which way, just asking a question for those who know the sector better.
Trying to update the government on my father's change of address for his state pension. The DWP phone line is clearly a creation of a disturbed sick mind.
On hold for 40mins, then a woman told me that i needed another another number (for carer's allowance), she was simply wrong, I expressed all I wanted to do was update Dad's address and then she talked over-me and hang up. Clearly, trying to improve stats. Called back, another 40min hold, then some other chap took my details for some yet unknown person to call me back in some dim and distant future, despite clearly having the system open there and then.
Now on hold for the HMRC. Slightly better since they are taking money in, but good grief.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
I don't think either Philip or myself were suggesting we build an entirely new company, more that it may make sense to encourage TSMC, Intel or indeed both to build a fab or 2 in the UK.
If it's bleeding edge, that's Sagans of money and a lot of time.
If it's older tech, that less money but no chance of profitability.
Just hope China doesn't do anything silly with Taiwan.
My day job is as a semis / electronics industry analyst. I've never seen a shortage like this one in over 30 years because it is so multi-faceted.
On the supply side it's not just leading edge foundry capacity that is tight; 8 inch legacy capacity is also fully booked. And in packaging, assembly and test, as well as tight capacity, there are also shortages of raw materials (lead frames, PCBs for advanced packaging, etc). We've also had the Texas freeze shutting down Samsung, Infineon and NXP fabs for a month and a fire at Renesas in Japan also shutting it down for a few weeks.
On the demand side, in addition to covid-related demand from hyperscalers (cloud servers and storage) and work/study from home PCs/Chromebooks, we also have the rapid migration to 5G smartphones (>500m units this year) and the rebound in automotive (including the shift to EV). 5G and EV need LOTS more chips per system for functions like RF and Power Management.
So, this situation isn't getting better any time soon - despite increased CAPEX now, it's takes a couple of years to bring a new fab up to full production. Wafer and chip ASPs are up significantly in 1Q and lead-times are out to a year for some chips. Double and triple ordering is rife. This situation will continue through 2021 and get gradually eased in 2022. The semis market will grow 20% this year, much of the revenue growth fueled by chip price rises.
The US is seriously worried (belatedly) about the national security implications of all this - i.e. the reliance on TSMC in Taiwan to build the world's semiconductors and the ambitions of China to develop it's own or "acquire" the technology it needs to build a competing technology roadmap over the next decade or two. EUV is the jewel in the crown here - without that, China will struggle to go head to head with TSMC, Samsung/SK hynix and Intel.
Hence the money being showered on Intel and TSMC by Biden to build fabs in Arizona ... this is about national security ("techo-nationalism" if you like). The globalized supply chain is going to be re-engineered, at least to some degree.
The interesting play here is the EU ... which way does it jump? Does it align with the US camp or cosy up to China? ASML (the only EUV supplier) is of course a European company ...
I've had some shares in ASML for a while, doing very well. They uniquely make EUV machines. I'm convinced someone will try to take them over.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
I don't think either Philip or myself were suggesting we build an entirely new company, more that it may make sense to encourage TSMC, Intel or indeed both to build a fab or 2 in the UK.
If it's bleeding edge, that's Sagans of money and a lot of time.
If it's older tech, that less money but no chance of profitability.
There is enough high margin demand left for 10nm and 14nm to justify new investment, but it doesn't make sense to do it here where, as I've said, it would be a standing start. However with Intel about to make the big switch to their 10nm and with them now renting out fab capacity that might be pretty fleeting as Intel's 14nm is pretty competitive with everyone else's 10nm processes.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
I don't think either Philip or myself were suggesting we build an entirely new company, more that it may make sense to encourage TSMC, Intel or indeed both to build a fab or 2 in the UK.
Intel maybe, but TSMC is highly unlikely to build outside of Taiwan. We'd probably be better off incentivising Global Foundries to build a 10nm fabrication plant here, it would be something covered by our agreement with the EU for tariff/quota free exports, and benefit from our entry to the CPTPP as well.
Another factor - for Taiwan, this kind of thing is a guarantee that the rest of the world won't sell them down the river, when the Chinese decide they want Taiwan back...
So Taiwan will be very keen to keep its position on the chip foundry business.
To be honest I doubt that's much of a factor. More important is [a] incentives, grants, tax breaks, etc [b] access to global supply chain and [c] existing infrastructure ... oh, and [d] long term commitment to provide ongoing support and stability.
I suppose when you've tried and failed with all the other radicalisations, right wing culture warrior is what's left. Galloway's novel is going to be an absolute stinker, pound shop orator soundbites repeated page after page.
"They hate the Crown but live well on its half crown."
The power and profundity of that.
Next up, in some TV studio somewhere or other, will surely come, "Boris Johnson and Nicola Sturgeon are two cheeks of the same arse".
He walked into the (unionist) party like he was loving where he was at. His hat strategically dipped below one eye, he likes to pretend he's a cat. He had one eye in the mirror as he watched himself talk tat. And all the Scots dreamed that he'd be their saviour, he'd be their saviour, and ...
He's a pain. I bet he thinks this poem is about him.
I have mentioned this before, but a close friend of mine, young and astrologically minded, said in late 2019 that "something terrible is going to happen to the whole world early next year" (she wasn't alone in this, apparently lots of zodiac-botherers foresaw the same, there were ominous alignments)
I mention this now because just this morning I remembered a conversation I had with her a few months later, after the pandemic began - about a year ago.
She said she was depressed. I asked her why. She replied "because I see where this will end. Governments will solve this with high technology, and we will move to a surveillance state, all countries will do this, it will be dystopia." She then added, "And there will be vaccines, but they will turn out to be dangerous"
True story.
Her later predictions weren't based on astrology, but just her hunch. Quite uncanny, looking back. A little bit of Brrr
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
Trouble is that the new technologies will have to be integrated into the existing silicon infrastructure - and the front line manufacturers (TSMC, Samsung, Intel, IBM etc) are already pursuing them. (I think one of the current ideas is transition metal dichalcogenide monolayers...)
Quantum computing is an interesting avenue. And would certainly benefit from increased government research funding.
I missed the previous Sunak thread but on its topic I'm not seeing anything that makes me question my bet. I'm letting my bet ride to the end (in one part because I don't know how to lay it off without putting £5000 in cash into Betfair considering it wasn't a Betfair bet to begin with), but I don't see anything in those texts that would cause Sunak to be fired from being Chancellor. I don't see anything there that will be significant for the next leadership election with MPs nominating or the membership vote either. So I don't see how its a factor.
I suspect Sunak's fate will be far more tied to whether and how the UK economy bounces back from Covid than anything to do with texts with Dave.
Greensill didn't get the loans they were after - so I'm at a loss as to why it's even vaguely a story. Sunak's emails are the exact thing you do to move an issue elsewhere
Indeed. Former minister lobbies current minister isn’t exactly a story.
Company employing former minister gets preferential treatment would have been; but the preferential treatment they hoped to get by employing the former minister never arrived.
It’s not really a huge story in itself.
However, it isn’t just a “former minister”, it’s a former prime minister, who I naively assumed would be above such things.
Can’t he swan off to the middle east to advocate for peace or something?
Like you I feel that this is more than a little grubby for an ex PM and it does not reflect well on him, whether it worked or not.
Dave (pbuh) doesn't have a sleazy bone in his body, I know he's turned down several jobs/directorships with firms because he didn't think it would be proper.
This story is an example of Dave being nice and helpful to people he knows.
He turned down several directorships and took Greensill Capital? Did he not ask George?
George has had a hectic private life in recent years, getting divorced, getting a new fiancée, knocking her up.
Just to make it even more exciting, Thea Rogers is Chief Customer Experience Offer at Deliveroo, which had a bit of a disaster recently with their IPO.
People are asking whose bright idea it was to spam the users of the App and their email and encourage the customers to invest in Deliveroo.
There is not a single part of the Deliveroo / Just Eat or Uber Eats business model that makes any actual sense. It's a combination of doing everything and anything to pretend their delivery workers aren't employees attached to charging takeaways 30% to attract customers too lazy to pick up the phone.
Its not about laziness remotely. That's an insane suggestion and is like suggesting that Amazon's business is about attracting customs too lazy to pick up the phone. Or anyone else that has an online website.
The ability to browse dozens of restaurants, pick one, browse their menu and then place an order all on one site makes great sense.
But the traditional takeaway model runs their delivery service at a loss (or break even at best) to attract business. Deliveroo is attempting to dominate and own a loss making business.
It is utterly nuts. It shows how abysmally capital is being allocated at the moment.
There's no reason deliveries need to be lossmaking.
The UK's largest delivery company surely is Domino's and if you look at Domino's profits and share price over the past twenty years they have not done badly at all.
Dominos is a pizza company that also delivers the pizza.
The delivery part doesn't make domino's any money. The pizza makes the money.
Deliveroo is attempting to take the delivery part of that business. The bit that doesn't make money.
Yep! They’re taking not making money, and scaling it up.
Like Amazon did for years.
I remember all the talk about how Amazon was a flawed model because they were making losses or extremely moderate profits for years despite the share price surging. Then from ~2016 onwards profit lifted off dramatically.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
Not suggesting an entire industry from scratch, heck not even suggesting we actually do anything.
Just thinking out loud, the security implications of this seem concerning regarding China/Taiwan etc - plus then having what happened with vaccines etc over the past 12 months.
Obviously we don't want to on-shore everything, but having some form of something domestically made seems like it could be a better idea for economic and national security than eg Trident.
OTOH it could be a mammoth blackhole and a terrible idea.
I have no opinion which way, just asking a question for those who know the sector better.
The difference is that in vaccines we still had a lot of the underlying expertise and a massive biotech industry that just hadn't been focussed on vaccines. It was a bit of a sleeping giant. In semiconductors the UK is literally nowhere.
What we should be doing is putting in lots of investment incentives for development of graphene into a next gen semiconductor industry. It's a better bet than anything we'd spend catching up in for silicon semiconductors.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
The other problem you hit with this stuff is that it needs continual re-investment. Each time the process changes - pretty much new build....
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Well 80% of TSMC revenue comes from processes older than 7nm, the cameras in your phone are most likely made on 28nm processes in a Sony foundry in Japan. There's a lot of silicon that doesn't need to be bleeding edge (car sensors, for example) and that's where the government might want to have some kind of national champion, but l I don't think it's a good idea as we'd need to literally build up an industry from nothing. Even building a 14nm immersion lithography fabrication plant would be billions invested for a not very big and very low margin market.
Not suggesting an entire industry from scratch, heck not even suggesting we actually do anything.
Just thinking out loud, the security implications of this seem concerning regarding China/Taiwan etc - plus then having what happened with vaccines etc over the past 12 months.
Obviously we don't want to on-shore everything, but having some form of something domestically made seems like it could be a better idea for economic and national security than eg Trident.
OTOH it could be a mammoth blackhole and a terrible idea.
I have no opinion which way, just asking a question for those who know the sector better.
The difference is that in vaccines we still had a lot of the underlying expertise and a massive biotech industry that just hadn't been focussed on vaccines. It was a bit of a sleeping giant. In semiconductors the UK is literally nowhere.
What we should be doing is putting in lots of investment incentives for development of graphene into a next gen semiconductor industry. It's a better bet than anything we'd spend catching up in for silicon semiconductors.
Thanks, that makes sense. Been a good discussion.
Do you think graphene will have security implications in the future like chips and vaccines do? Or just economic ones?
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
Trouble is that the new technologies will have to be integrated into the existing silicon infrastructure - and the front line manufacturers (TSMC, Samsung, Intel, IBM etc) are already pursuing them. (I think one of the current ideas is transition metal dichalcogenide monolayers...)
Quantum computing is an interesting avenue. And would certainly benefit from increased government research funding.
To some degree, but there is likely going to be a gap during that transition where a graphene startup can scale quite quickly without needing to deal with legacy costs. The government should be looking to exploit that. Agreed on quantum computing where we are probably among the world leaders for it.
I suppose when you've tried and failed with all the other radicalisations, right wing culture warrior is what's left. Galloway's novel is going to be an absolute stinker, pound shop orator soundbites repeated page after page.
"They hate the Crown but live well on its half crown."
The power and profundity of that.
Next up, in some TV studio somewhere or other, will surely come, "Boris Johnson and Nicola Sturgeon are two cheeks of the same arse".
He walked into the (unionist) party like he was loving where he was at. His hat strategically dipped below one eye, he likes to pretend he's a cat. He had one eye in the mirror as he watched himself talk tat. And all the Scots dreamed that he'd be their saviour, he'd be their saviour, and ...
He's a pain. I bet he thinks this poem is about him.
"They hate the Crown but live well on its half crown" is a bloody brilliant line. Vivid, powerful, crisp, and encapsulates a great truth in one apt metaphor
Only problem is that it's so dated - half crowns?
He should be writing scripts about the Irish Rebellion, that would be a stand-out line in any screenplay
Minidisc sound quality is good - Sony ATRAC compression (especially type-S on later models) is very good compared to a lot of the lossy formats out there which are newer. Besides which I have tinnitus so I do not need audiophile 24bit quality.
ATRAC may not be the worst compression algorithm, but it axiomatically worse than uncompressed sources.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
Well, I'm biased of course but if you think about what will drive the economies of the future (5G/6G, A.I., IoT, Power and Energy, Automation, EV and Autonomous Vehicles and so on), semis of all flavours will be at the heart of those technologies.
So, in my view, semiconductor technology is absolutely a national security imperative. China knows this. The US forgot this but has finally woken up to it. The EU is not a good organization to think about this either strategically or operationally (just look at the vaccine shambles). If the UK is smart it will realize the national security implications and do something about it. It's not an easy sell though because it's not as in your face politically as "overwhelm the NHS " or "the climate emergency".
Even as someone who isn't a royalist, I have great respect for him, even for simply the Duke of Edinburgh scheme for teenagers which was a great thing he set up. Rest in peace.
Would it make sense for the UK to invest in this unilaterally?
How much would it cost to become a leading chipmaker domestically? And given the national security implications if we were to contrast that with eg the cost of Trident how would it compare?
Having a domestic supply of chips seems like it could be a national security issue and an infinitely more useful thing economically than Trident.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
I suppose when you've tried and failed with all the other radicalisations, right wing culture warrior is what's left. Galloway's novel is going to be an absolute stinker, pound shop orator soundbites repeated page after page.
"They hate the Crown but live well on its half crown."
The power and profundity of that.
Next up, in some TV studio somewhere or other, will surely come, "Boris Johnson and Nicola Sturgeon are two cheeks of the same arse".
He walked into the (unionist) party like he was loving where he was at. His hat strategically dipped below one eye, he likes to pretend he's a cat. He had one eye in the mirror as he watched himself talk tat. And all the Scots dreamed that he'd be their saviour, he'd be their saviour, and ...
He's a pain. I bet he thinks this poem is about him.
I've only just realised what was his main motivation for moving to the region of Dumfries and Galloway, aside from it being stuffed with credulous Unionists: Dumfries & Galloway, c'est moi.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
Well, I'm biased of course but if you think about what will drive the economies of the future (5G/6G, A.I., IoT, Power and Energy, Automation, EV and Autonomous Vehicles and so on), semis of all flavours will be at the heart of those technologies.
So, in my view, semiconductor technology is absolutely a national security imperative. China knows this. The US forgot this but has finally woken up to it. The EU is not a good organization to think about this either strategically or operationally (just look at the vaccine shambles). If the UK is smart it will realize the national security implications and do something about it. It's not an easy sell though because it's not as in your face politically as "overwhelm the NHS " or "the climate emergency".
If we want to do something about that in the UK, we need to think in terms of building up knowledge and participation in the supply chains first.
Simply announcing Mao style, that we will have a bleeding edge chip fab isn't going to work.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
Yes, I agree. No point in competing on silicon. Much better to target new materials. But you'd still need to ramp up and scale a similar type of manufacturing expertise and silicon would be a way to do that.
I suppose when you've tried and failed with all the other radicalisations, right wing culture warrior is what's left. Galloway's novel is going to be an absolute stinker, pound shop orator soundbites repeated page after page.
"They hate the Crown but live well on its half crown."
The power and profundity of that.
Next up, in some TV studio somewhere or other, will surely come, "Boris Johnson and Nicola Sturgeon are two cheeks of the same arse".
He walked into the (unionist) party like he was loving where he was at. His hat strategically dipped below one eye, he likes to pretend he's a cat. He had one eye in the mirror as he watched himself talk tat. And all the Scots dreamed that he'd be their saviour, he'd be their saviour, and ...
He's a pain. I bet he thinks this poem is about him.
"They hate the Crown but live well on its half crown" is a bloody brilliant line. Vivid, powerful, crisp, and encapsulates a great truth in one apt metaphor
Only problem is that it's so dated - half crowns?
He should be writing scripts about the Irish Rebellion, that would be a stand-out line in any screenplay
Mmm. I suppose.
Hey, so scrolling through last night's thread I see that you led a rich exchange about how (post 'metoo') the pendulum has perhaps now swung too far against sexual harassment.
Sorry to have missed that. Could have added value.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Yes, starting from scratch we'd need to push at least £10bn in the next year for literally no return and then double that next year for still almost no return. It wouldn't be until year 3 and another £20-25bn until we started to see 5nm chips, by which time TSMC would be on a better process node.
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
Do you think it would be worth doing that, potentially using some money from 'national security' budgets and not purely economic? Or not really?
I think if we were to do something it should be in next generation technology like graphene. A standing start for a silicon semiconductor industry vs TSMC and Intel just seems completely unrealistic and a huge, huge money blackhole for the government which will get nothing for it but hassle and bad headlines of money wasted it could have spent on the NHS etc...
Yes, I agree. No point in competing on silicon. Much better to target new materials. But you'd still need to ramp up and scale a similar type of manufacturing expertise and silicon would be a way to do that.
A thread header on what we might do for the cost of HS2, Trident or a bridge/tunnel to Ireland... ?
5...4...3...2...1....somebody sends an offensive tweet about it.
They're going to replay his greatest hits? Slitty eyes, always drunk etc?
I suspect he would have appreciated that.
He said some bloody amazing things, but always meant well. Just a product of his era. The Duke of Edinburgh's Award scheme will continue to honour his memory.
Thoughts with Her Majesty, they were married for 73 years, and known each other for longer, that's a testament to their love.
It's a known medical phenomenon that with old married couples, who have spent a looooong life together, that if one keels over the next follows shortly after. Hope I'm wrong, obvs
Sad news about the Duke. Nearly met him once. It was at a function in the Long Room at Lords. The Duke was there, sipping his G&T, when he made eye contact with our group and headed over. Alas, he was distracted by some attractive young ladies along the way and never made it.
Would it make sense for the UK to invest in this unilaterally?
How much would it cost to become a leading chipmaker domestically? And given the national security implications if we were to contrast that with eg the cost of Trident how would it compare?
Having a domestic supply of chips seems like it could be a national security issue and an infinitely more useful thing economically than Trident.
If the UK wanted to be serious it would need to commit to a long term investment in semis i.e. decades. You can't do this half-heartedly. $10B for a fab is the tip of the iceberg and these things can't be done standalone - you need all the surrounding infrastructure, eco-systems, R&D, etc, etc.). The UK could do worse than look at what South Korea and Taiwan did in the 1980s and 1990s to kick start their semis industries.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
it wouldn't be as easy as S Korea or Taiwan found it a few decades back, either. Taiwan more or less invented the pure foundry model, which meant they got ever smart engineer form the US who wanted to start their own company turning up as a customer. We don't even have much a domestic market to support that, and nothing to attract overseas customers. Really difficult to see how we could do it without attracting major established manufacturers, I think ?
That said, it might still be a better use of money than Trident/Trident replacement (which might end up getting scrapped anyway should it lose its Scottish base).
Yes, you'd need to attract an established manufacturer - not impossible but highly unlikely I would say.
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
Something not generally realised - in the pre-radar (and early radar) days, searchlights were extremely important. They were used to blind the enemy and illuminate them for the rangefinders. If you got it wrong, you were providing a perfect target for the enemy.
The RN learned this the hard way, in WWI.
Which is why Philip, as a young officer, got a Mention in Dispatches from Cunningham. Who didn't give out praise to just anyone...
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
Most importantly, will Harry and Meghan come over for the funeral? We must know.
They HAVE to. Harry would become the most hated man in Britain if he refused to attend the funeral of his late grandfather, and snubbed his grandmother, even in her grief
Thoughts with Her Majesty, they were married for 73 years, and known each other for longer, that's a testament to their love.
It's a known medical phenomenon that with old married couples, who have spent a looooong life together, that if one keels over the next follows shortly after. Hope I'm wrong, obvs
That happened to my nan and grandad. My nan died after a long illness, but grandad was fit as a fiddle...he died in his sleep a month later.
Just hope China doesn't do anything silly with Taiwan.
My day job is as a semis / electronics industry analyst. I've never seen a shortage like this one in over 30 years because it is so multi-faceted.
On the supply side it's not just leading edge foundry capacity that is tight; 8 inch legacy capacity is also fully booked. And in packaging, assembly and test, as well as tight capacity, there are also shortages of raw materials (lead frames, PCBs for advanced packaging, etc). We've also had the Texas freeze shutting down Samsung, Infineon and NXP fabs for a month and a fire at Renesas in Japan also shutting it down for a few weeks.
On the demand side, in addition to covid-related demand from hyperscalers (cloud servers and storage) and work/study from home PCs/Chromebooks, we also have the rapid migration to 5G smartphones (>500m units this year) and the rebound in automotive (including the shift to EV). 5G and EV need LOTS more chips per system for functions like RF and Power Management.
So, this situation isn't getting better any time soon - despite increased CAPEX now, it's takes a couple of years to bring a new fab up to full production. Wafer and chip ASPs are up significantly in 1Q and lead-times are out to a year for some chips. Double and triple ordering is rife. This situation will continue through 2021 and get gradually eased in 2022. The semis market will grow 20% this year, much of the revenue growth fueled by chip price rises.
The US is seriously worried (belatedly) about the national security implications of all this - i.e. the reliance on TSMC in Taiwan to build the world's semiconductors and the ambitions of China to develop it's own or "acquire" the technology it needs to build a competing technology roadmap over the next decade or two. EUV is the jewel in the crown here - without that, China will struggle to go head to head with TSMC, Samsung/SK hynix and Intel.
Hence the money being showered on Intel and TSMC by Biden to build fabs in Arizona ... this is about national security ("techo-nationalism" if you like). The globalized supply chain is going to be re-engineered, at least to some degree.
The interesting play here is the EU ... which way does it jump? Does it align with the US camp or cosy up to China? ASML (the only EUV supplier) is of course a European company ...
Excellent post, and yet more proof that there is an expert in nearly everything hanging around PB.
Hopefully an opportunity for the UK, as large companies look at how the EU are treating pharmaceutical companies in an emergency. Would you out your $2bn of capital into a place where your output is threatened with confiscation when it’s most needed?
Agreed - it's great to get comments form people who actually know what they're taking about (as opposed to me).
I'm not sure how much of an opportunity it might be for the UK, though. Key equipment suppliers (Dutch ASML) are, as @ridaligo points out European - and Intel already has a very large presence in S. Ireland. We have minimal manufacturing in any part of the supply chain, so it would take some pretty big incentives to get any of the large manufacturers to locate here.
As for political supply risk, Taiwan is a great deal worse than the EU...
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
Full state funeral I assume?
Operation Forth Bridge. Interesting to see how the current restrictions are applied...
Thoughts with Her Majesty, they were married for 73 years, and known each other for longer, that's a testament to their love.
It's a known medical phenomenon that with old married couples, who have spent a looooong life together, that if one keels over the next follows shortly after. Hope I'm wrong, obvs
Likely. Charles needs to pass on the gig and give it to William.
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
You are allowed to break quarantine rules for a funeral, they are exempted
Most importantly, will Harry and Meghan come over for the funeral? We must know.
They HAVE to. Harry would become the most hated man in Britain if he refused to attend the funeral of his late grandfather, and snubbed his grandmother, even in her grief
They have no choice
You’re assuming the Royals would want to invite them in the first place. Why would they want the outcasts to make a family funeral all about themselves?
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
IIRC travel from the US to the UK is allowed at the moment. The problem would be with the return to the US, I think.
Thoughts with Her Majesty, they were married for 73 years, and known each other for longer, that's a testament to their love.
It's a known medical phenomenon that with old married couples, who have spent a looooong life together, that if one keels over the next follows shortly after. Hope I'm wrong, obvs
That happened to my nan and grandad. My nan died after a long illness, but grandad was fit as a fiddle...he died in his sleep a month later.
Yes, it is a sad yet touching behaviourism. I have a sense it has been observed in the animal kingdom, too
Most importantly, will Harry and Meghan come over for the funeral? We must know.
How can they, given quarantine rules? I am assuming it will be a private funeral with a memorial service later when restrictions have been lifted, as well.
Full state funeral I assume?
Well, how can it be under current restrictions? That’s why I think there might be some flexibility demanded.
All other considerations aside, even if Her Majesty has been double-jabbed, they won’t want Covid to potentially get into the royal household.
Most importantly, will Harry and Meghan come over for the funeral? We must know.
They HAVE to. Harry would become the most hated man in Britain if he refused to attend the funeral of his late grandfather, and snubbed his grandmother, even in her grief
They have no choice
Will they have to do 14 days in a Premier Inn quarantine? That would be quite amusing.
Comments
There's a lot of money to be made in the lesser nodes which also requires separate investment but to a lesser degree but that would put UKSMC in competition with Intel, TSMC, GloFo and a whole host of Japanese foundries.
You'd hope the UK would be looking at it because we have so many in-built advantages (political stability, no natural disasters, English-speaking, good Universities, good global trade links, etc, etc) but I suspect the focus is more on life-sciences and green tech.
A federal union may work. But I don't see anyone other than the LibDems seriously arguing for it. So England will try to colonially repress the savages in the provinces, but ultimately that only drives harder the forces tearing the union apart.
If England wants supremacy to do what it wants without regard to what the people it trades with wants, then sadly we should let it.
But it’s Glaws. They’ll find inventive ways to combust if given a little time.
https://twitter.com/GerryHassan/status/1380460955612635136?s=20
So figure on multiple billions per year, continuously. You should be building the next fab while running the old one(s).
This kind of rapid obsolescence figured in Teslas early success. A similar, but lest expensive, obsolescence occurs in lithium battery manufacture. Last years plant is of little interest.... What Tesla did was to realise that (a) nearly all battery problems are in the bleeding edge, newest battery designs, (b) last years battery designs are available for low low prices.
So they did a deal with Panasonic - they would buy 100% of the production from the just-obsolete plants. For a very low price....
Wasn't our chip-maker ...... somewhere in Cambridge IIRC ...... sold off some time ago?
ARM are a chip/architecture designer I think, rather than a chip-maker.
I could be wrong.
You're not wrong. ARM is an IP provider; it doesn't make anything.
Taiwan more or less invented the pure foundry model, which meant they got ever smart engineer form the US who wanted to start their own company turning up as a customer. We don't even have much a domestic market to support that, and nothing to attract overseas customers.
Really difficult to see how we could do it without attracting major established manufacturers, I think ?
That said, it might still be a better use of money than Trident/Trident replacement (which might end up getting scrapped anyway should it lose its Scottish base).
So I don't think it would ever turn a profit.
Just thinking out loud, the security implications of this seem concerning regarding China/Taiwan etc - plus then having what happened with vaccines etc over the past 12 months.
Obviously we don't want to on-shore everything, but having some form of something domestically made seems like it could be a better idea for economic and national security than eg Trident.
OTOH it could be a mammoth blackhole and a terrible idea.
I have no opinion which way, just asking a question for those who know the sector better.
Trying to update the government on my father's change of address for his state pension. The DWP phone line is clearly a creation of a disturbed sick mind.
On hold for 40mins, then a woman told me that i needed another another number (for carer's allowance), she was simply wrong, I expressed all I wanted to do was update Dad's address and then she talked over-me and hang up. Clearly, trying to improve stats. Called back, another 40min hold, then some other chap took my details for some yet unknown person to call me back in some dim and distant future, despite clearly having the system open there and then.
Now on hold for the HMRC. Slightly better since they are taking money in, but good grief.
If it's older tech, that less money but no chance of profitability.
Remember the Betfred "Fruit machine" winner who won £1.7m and Betfred refused to pay out - well they've lost the court case https://www.theregister.com/2021/04/08/betfred_uk_payout/
So Taiwan will be very keen to keep its position on the chip foundry business.
https://twitter.com/paul_lever/status/1380457245675286535?s=21
All 3 parties caused the problems- all 3 own finding solutions
If we hadn't left the EU, of course .......
To be honest I doubt that's much of a factor. More important is [a] incentives, grants, tax breaks, etc [b] access to global supply chain and [c] existing infrastructure ... oh, and [d] long term commitment to provide ongoing support and stability.
The power and profundity of that.
Next up, in some TV studio somewhere or other, will surely come, "Boris Johnson and Nicola Sturgeon are two cheeks of the same arse".
He walked into the (unionist) party like he was loving where he was at.
His hat strategically dipped below one eye, he likes to pretend he's a cat.
He had one eye in the mirror as he watched himself talk tat.
And all the Scots dreamed that he'd be their saviour, he'd be their saviour, and ...
He's a pain.
I bet he thinks this poem is about him.
I have mentioned this before, but a close friend of mine, young and astrologically minded, said in late 2019 that "something terrible is going to happen to the whole world early next year" (she wasn't alone in this, apparently lots of zodiac-botherers foresaw the same, there were ominous alignments)
I mention this now because just this morning I remembered a conversation I had with her a few months later, after the pandemic began - about a year ago.
She said she was depressed. I asked her why. She replied "because I see where this will end. Governments will solve this with high technology, and we will move to a surveillance state, all countries will do this, it will be dystopia." She then added, "And there will be vaccines, but they will turn out to be dangerous"
True story.
Her later predictions weren't based on astrology, but just her hunch. Quite uncanny, looking back. A little bit of Brrr
(I think one of the current ideas is transition metal dichalcogenide monolayers...)
Quantum computing is an interesting avenue.
And would certainly benefit from increased government research funding.
What we should be doing is putting in lots of investment incentives for development of graphene into a next gen semiconductor industry. It's a better bet than anything we'd spend catching up in for silicon semiconductors.
Do you think graphene will have security implications in the future like chips and vaccines do? Or just economic ones?
Only problem is that it's so dated - half crowns?
He should be writing scripts about the Irish Rebellion, that would be a stand-out line in any screenplay
So, in my view, semiconductor technology is absolutely a national security imperative. China knows this. The US forgot this but has finally woken up to it. The EU is not a good organization to think about this either strategically or operationally (just look at the vaccine shambles). If the UK is smart it will realize the national security implications and do something about it. It's not an easy sell though because it's not as in your face politically as "overwhelm the NHS " or "the climate emergency".
Even as someone who isn't a royalist, I have great respect for him, even for simply the Duke of Edinburgh scheme for teenagers which was a great thing he set up. Rest in peace.
RIP.
TSMC recently suggested $100B CAPEX and R&D over the next 3 years, which is probably on the high side for a short term investment plan (but they know what they are doing).
For a serious, long term UK program, start with a fund of $100B for the first 5-year plan and go from there.
Programme.
Interesting post thought.
This will dominate the media for days rightly or wrongly
Simply announcing Mao style, that we will have a bleeding edge chip fab isn't going to work.
Hey, so scrolling through last night's thread I see that you led a rich exchange about how (post 'metoo') the pendulum has perhaps now swung too far against sexual harassment.
Sorry to have missed that. Could have added value.
RIP The Duke after a lifetime of service
That’s going to leave an enormous void for her, although it doesn’t look as though it’s come as a great shock.
I hope the government have drawn up plans for a Regency in case she is taken ill.
https://en.wikipedia.org/wiki/Battle_of_Cape_Matapan#Night_action
Something not generally realised - in the pre-radar (and early radar) days, searchlights were extremely important. They were used to blind the enemy and illuminate them for the rangefinders. If you got it wrong, you were providing a perfect target for the enemy.
The RN learned this the hard way, in WWI.
Which is why Philip, as a young officer, got a Mention in Dispatches from Cunningham. Who didn't give out praise to just anyone...
They have no choice
All other considerations aside, even if Her Majesty has been double-jabbed, they won’t want Covid to potentially get into the royal household.