I think this company was set up in partnership with IBM, so maybe they can solve the issue of EUV machines, but not the rivers of money needed. I don't think the Japanese government in its current state is going to hand over tens of billions...
Rapidus is financed by Denso, Kioxia, MUFG Bank, NEC, NTT, SoftBank, Sony, and Toyota. And the Japanese government is also financing it via the Ministry of Economy Trade and Industry. I reckon they have the money
Capital flows to where it'll create the best return. There are trillions waiting to be invested. I imagine a portion of that will 'flow' to ventures like this one.
Talent on the other hand... I imagine we'll be seeing LOTS more robots and automation.
Let's hope they succeed, TSMC is too crowded and hopefully another player can bring the prices down, but the ASML production bottleneck gets even tighter.
In my humble opinion, governments are gunning for their own domestic fabs because of what happened to Huawei's business when Trump decided to throw a dynamite in it.
But they're about to find out just how hard it is to do cutting edge chip manufacturing at scale with good enough yields that can actually generate a profit. Even Intel, with its decades of experience, has failed in the last few years. Samsung has seemingly failed as well. I have a hard time believing that a brand new venture will be able to produce 2nm/1.4nm chips effectively that isn't making a loss.
The only country that I think can actually pull this off is China because of their immense will, talent, and lower cost of labor.
I think your observations are fair, but perhaps you're overstating the difficulty of initiating manufacture. The US people can prove surprisingly creative when they feel endangered or if someone is able to unify them against a common threat.
China, on the other hand, has a more societal and collective benefit focus. They've also been at this effort for a bit to break dependence from potential geopolitical threats so they've a fair bit of momentum.
Both will likely lose the synergies of working together, but ultimately be successful.
I think the motivation is that everybody has become reliant on TSMC, and that TSMC is located on an island that China might invade within five years. So for national security reasons people want cutting edge domestic fabs. It has little to do with Trump.
The key to profitability in semiconductors isn't a bunch of work ethic / worker suffering /cultural blah blah blah. It's doubt having enough demand to support economies of scale. If you've got that, and some minimum level of patience, capitalization, and management competence, then you can do it.
Thirty years ago, the PC was the driver of demand and economies of scale, and Intel flourished because of their position in the PC market. In the 2010s, the smartphone became the demand driver, and TSMC flourished because Intel dropped the ball on mobile, allowing Intel to pick it up.
Today, the PC and mobile markets are still with us, but we are adding a bunch for AI, which might in turn rejuvenate the other two markets. We might also see some robotics/drones around the corner.
So it's quite possible that there will be plenty of demand to support all of these players.
I think its both. Trump showed the world what happens when you are cut off from TSMC and that spooked, well, everyone when they realized that ALL of TSMC at the time was on one singular island. Forget china, imagine if one big tsunami hit them.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
12 Comments
Back to Article
Terry_Craig - Friday, May 24, 2024 - link
For real... Where do they think this money will come from?ballsystemlord - Friday, May 24, 2024 - link
Where will the talent come from?Where will the EUV machines come from? IIRC, ASML is booked up through 2025.
Dante Verizon - Friday, May 24, 2024 - link
I think this company was set up in partnership with IBM, so maybe they can solve the issue of EUV machines, but not the rivers of money needed. I don't think the Japanese government in its current state is going to hand over tens of billions...PeachNCream - Sunday, May 26, 2024 - link
Considering that they won't begin production until 2027 per the article, ASML's current production allocation is probably a non-factor.Threska - Friday, May 24, 2024 - link
With high-NA EUV one can get some very nice $1000 printing plates.gdansk - Monday, May 27, 2024 - link
Rapidus is financed by Denso, Kioxia, MUFG Bank, NEC, NTT, SoftBank, Sony, and Toyota. And the Japanese government is also financing it via the Ministry of Economy Trade and Industry.I reckon they have the money
gteichrow - Tuesday, May 28, 2024 - link
Capital flows to where it'll create the best return. There are trillions waiting to be invested. I imagine a portion of that will 'flow' to ventures like this one.Talent on the other hand... I imagine we'll be seeing LOTS more robots and automation.
Zoolook13 - Saturday, May 25, 2024 - link
Let's hope they succeed, TSMC is too crowded and hopefully another player can bring the prices down, but the ASML production bottleneck gets even tighter.lemurbutton - Monday, May 27, 2024 - link
In my humble opinion, governments are gunning for their own domestic fabs because of what happened to Huawei's business when Trump decided to throw a dynamite in it.But they're about to find out just how hard it is to do cutting edge chip manufacturing at scale with good enough yields that can actually generate a profit. Even Intel, with its decades of experience, has failed in the last few years. Samsung has seemingly failed as well. I have a hard time believing that a brand new venture will be able to produce 2nm/1.4nm chips effectively that isn't making a loss.
The only country that I think can actually pull this off is China because of their immense will, talent, and lower cost of labor.
PeachNCream - Monday, May 27, 2024 - link
I think your observations are fair, but perhaps you're overstating the difficulty of initiating manufacture. The US people can prove surprisingly creative when they feel endangered or if someone is able to unify them against a common threat.China, on the other hand, has a more societal and collective benefit focus. They've also been at this effort for a bit to break dependence from potential geopolitical threats so they've a fair bit of momentum.
Both will likely lose the synergies of working together, but ultimately be successful.
Blastdoor - Tuesday, May 28, 2024 - link
I think the motivation is that everybody has become reliant on TSMC, and that TSMC is located on an island that China might invade within five years. So for national security reasons people want cutting edge domestic fabs. It has little to do with Trump.The key to profitability in semiconductors isn't a bunch of work ethic / worker suffering /cultural blah blah blah. It's doubt having enough demand to support economies of scale. If you've got that, and some minimum level of patience, capitalization, and management competence, then you can do it.
Thirty years ago, the PC was the driver of demand and economies of scale, and Intel flourished because of their position in the PC market. In the 2010s, the smartphone became the demand driver, and TSMC flourished because Intel dropped the ball on mobile, allowing Intel to pick it up.
Today, the PC and mobile markets are still with us, but we are adding a bunch for AI, which might in turn rejuvenate the other two markets. We might also see some robotics/drones around the corner.
So it's quite possible that there will be plenty of demand to support all of these players.
TheinsanegamerN - Wednesday, May 29, 2024 - link
I think its both. Trump showed the world what happens when you are cut off from TSMC and that spooked, well, everyone when they realized that ALL of TSMC at the time was on one singular island. Forget china, imagine if one big tsunami hit them.