If you look at the next-generation lithography road maps, even in the early 90s EUV was just beyond the horizon, but it was on the road map. By 2014 more than twenty billion dollars were spent by various governments and private research when ASML began to ship these https://web.archive.org/web/20150920221539/http://www.asml.c... and even so it was another four years before the first actual devices using EUV CPUs shipped. And, let's not forget, the first device was the A12 Bionic for the iPhone XS/XS Max/XR and some iPads. In the fourth quarter of 2018 alone, Apple sold 46.89 million iPhones worldwide. And there were other companies jumping on the bandwagon but that's a lot of preorder money to help bootstrapping this.
Taiwanese engineers lured to mainland China as chip makers go into overdrive
"More than 300 senior engineers from Taiwan have moved to mainland chip makers so far this year, joining nearly 1,000 others who have relocated since Beijing set up a US$22 billion fund to develop the chip industry in 2014"
Even with the "chokepoints" it is just a question of time and money, no? In particular now that China essentially is being forced to be self-sufficient.
The hardest thing about producing the atomic bomb itself, was to determine if it was actually possible. But once it was achieved, then everyone knew it was possible. So all you had to do was throw more time and money into cracking it. And a little espionage probably didn’t hurt too.
The 5nm EUV, while it sounds cutting edge today, may just need enough time, money, and focus to achieve it.
Someone already achieved it, so we know it is possible. And SMIC may go a different route to achieve the same result, thus giving China independent Intellectual Property rights to it.
Not sure whether we are talking of the same problem but the Oak Ridge site alone cost more than a billion dollars and nothing cost a billion those years (the total federal budget outlay in 1942 was a modest 35 billion which admittedly climbed to almost 93 billion by 1945). The K-25 building was half of that: 1,640,000 square feet floor space, 97,500,000 cubic feet volume. The Hanford site was another 400 million. They created an entire industry out of nothing.
EV is not that bad but yes, it's quite bad.
And the possibility was always known, the devil really is in the details.
Still, the Soviet Union somehow repeated that. Despite the fact that half the country was still in ruins. Kiev and Minsk don't have that many pre-war buildings.
What? No, the hardest thing about producing an atomic bomb is spinning up enough U235. To produce highly enriched uranium requires huge numbers of gas centrifuges. This makes your facilities very large, conspicuous targets for espionage and sabotage.
If the hardest part was knowing nuclear weapons are possible then the world would be a giant glass parking lot by now.
> the hardest thing about producing an atomic bomb is spinning up enough U235. To produce highly enriched uranium requires huge numbers of gas centrifuges.
This is true in the modern era. However it should be noted that the Manhattan Project did not use gas centrifuges to produce their enriched uranium. Centrifuges were apparently considered but ultimately gaseous diffusion was used. Gaseous diffusion requires larger facilities and more energy, but I suppose the process doesn't require such tight tolerances as modern highspeed centrifuges.
If you look at the next-generation lithography road maps, even in the early 90s EUV was just beyond the horizon, but it was on the road map. By 2014 more than twenty billion dollars were spent by various governments and private research when ASML began to ship these https://web.archive.org/web/20150920221539/http://www.asml.c... and even so it was another four years before the first actual devices using EUV CPUs shipped. And, let's not forget, the first device was the A12 Bionic for the iPhone XS/XS Max/XR and some iPads. In the fourth quarter of 2018 alone, Apple sold 46.89 million iPhones worldwide. And there were other companies jumping on the bandwagon but that's a lot of preorder money to help bootstrapping this.