The race to lead in artificial intelligence isn't the only event in which the U.S. and China are competing for dominance. The pursuit of fusion — the "Holy Grail" of clean energy — is also pitting the superpowers against each other, and American tech leaders worry China could surge ahead.
At a Technology Alliance conference on Tuesday, Washington state companies building commercial fusion technologies raised concerns about China's strategy to pour resources into fusion.
"The U.S. is not committed to fusion. China is, by orders of magnitude," said Ben Levitt, the head of R&D for Zap Energy, speaking on a fusion panel at the Seattle Investor Summit+Showcase.
While the U.S. government spent approximately $800 million a year on fusion efforts during the Biden administration, China is investing more than twice that annually, IEEE Spectrum and others report. The Trump administration has taken action supporting nuclear fission, which powers today's nuclear reactors, but has not shown the same interest in fusion. The sector has become increasingly reliant on venture capital to fund its progress.
China is also focused on training fusion physicists and engineers, while President Trump is slashing funding for scientific research.
Fusion is so highly sought after given its potential to provide nearly limitless, carbon-free power, which could be critical to meet growing energy demands from AI applications and the global push to decarbonize transportation, the electrical grid, heating and cooling, industrial applications and elsewhere.
"The U.S. started with a very good hand in fusion and has played it extremely poorly," Levitt said. "So, yeah, we're definitely on the back foot."
The conference panel also included Brian Riordan, co-founder and chief operating officer of Avalanche Energy, and Anthony Pancotti, co-founder and head of R&D for Helion Energy.
Riordan argued that while China appears to be making strides in the race, what matters even more is who develops the most affordable technology.
Physicists for decades have pursued fusion energy. But replicating the reactions that power the Sun and stars is massively challenging and requires technologies that can generate super high pressure and temperatures of 100 million degrees Celsius, and sustain those conditions — plus efficiently capture the energy that fusion produces.
In December 2022, the U.S. National Ignition Facility (NIF) at Lawrence Livermore National Laboratory hit a key milestone in fusion research, demonstrating that fusion reactions here on Earth could release more power than required to produce them.
Images published in January revealed that China appears to be building a fusion research facility modeled on NIF — but even larger. Others suggest the site could be a giant Z-pinch machine — similar to the technology being pursued by Zap.
Years ago, a Chinese website posted a graphic of a fusion device that bore a troubling resemblance to Helion's technology, the company has said.
"We have seen copycats in China already, and it is terrifying," Pancotti said on Tuesday. "They can mobilize people and money at a scale that is beyond even what venture capital can do in this country. And so I think there's real concern there, and there's real concern around supply chain, too."
Added Levitt: "I wouldn't be surprised if every single one of our [fusion] concepts has a city designated to it in China."
When it comes to world ending tech, I'm not sure I want it to be a race.
(Score: 3, Insightful) by khallow on Wednesday June 11, @11:07AM (21 children)
(Score: 2) by PiMuNu on Wednesday June 11, @11:10AM (2 children)
On a technical note, HTS superconductors are a game-changer in this field and are now a reality.
(Score: 1) by khallow on Wednesday June 11, @11:23AM (1 child)
(Score: 5, Interesting) by PiMuNu on Wednesday June 11, @11:44AM
> They've been a reality in fusion power for at least 20 years.
I don't think that is true, certainly not with the fields that are currently being touted. For example the highest field commercial solenoid is Bruker's 28 T magnet, which only became available a couple of years ago. These sorts of performances have only been possible because mitigation measures for coil delamination of the HTS ribbons have only recently been figured out.
(Score: 3, Insightful) by c0lo on Wednesday June 11, @11:45AM (17 children)
Ummm... like what institutional obstructions?
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 2, Interesting) by Anonymous Coward on Wednesday June 11, @04:28PM (2 children)
Big Oil owns half the politicians, Big Green owns the other half. Both hate and fear Big Nuke, and they've bought half a century of regulation devoted to making nuclear unviable and unable to compete with their heavily-subsidized and externality-hidden industries. Nowadays the fuel has to be bought from Russia because the regulatory environment in the US is highly prohibitive. How is that going to play politically?
(Score: 3, Funny) by epitaxial on Wednesday June 11, @04:59PM
Bullshit. All "big oil" has to do is license this fictional technology they're holding back.
(Score: 4, Informative) by c0lo on Wednesday June 11, @07:47PM
If there are so many regulations on nuke fusion, one thinks you would be able to cite/link at least one.
You mean deuterium and hydrogen?
Look, deuterium oxide for $249/100g [unitednuclear.com] for housewives or anyone w/ money to pay. Put it in an electrolyser, get you deuterium.
Scroll down the page and read the "quick check". Like... huh, no restrictions.
Or buy it by the drum from Canada [isowater.com].
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @04:40AM (5 children)
(Score: 2) by c0lo on Thursday June 12, @04:59AM (4 children)
And the willingness to pursue a certain approach is a(n institutional) obstruction to other approaches... exactly how?
You sayin' Helion, SPARC, Commonwealth Fusion Systems, TAE Technologies (and god knows who others) were somehow obstructed by the US govt/institutions/regulation? if so, can you provide citations?
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @06:47AM (3 children)
The problem with these big projects is that they take the resources. Government funding is so often a zero sum game. So when a huge project with little future grabs all the funding? Well, that doesn't leave much for the rest.
(Score: 2) by c0lo on Thursday June 12, @07:51AM (2 children)
Maybe, but that's not an obstruction, much less a regulatory one. At best it's a limitation and there will always be projects that will be subjected to such limits.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @12:03PM (1 child)
Depends on your metric. If you're trying to build a commercial power plant, then money spent on white elephants like ITER isn't being spent on you.
A limitation that obstructs. Not seeing the need for nuance here. It's one thing to be limited because you're competing for funding with other fusion power approaches that have a reasonable chance of becoming commercial. And another to be competing with systems that don't have a ghost of a chance of becoming commercial. ITER falls in that latter category.
(Score: 2) by c0lo on Thursday June 12, @02:02PM
(sterile conversation thread)
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @04:56AM (7 children)
(Score: 3, Interesting) by c0lo on Thursday June 12, @05:37AM (6 children)
This if how the first working transistor looked like in 1947 [computerhistory.org] - didn't look too commercial at the time either.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @06:20AM (5 children)
(Score: 1) by khallow on Thursday June 12, @06:59AM
I'm starting to think this might be a Wikipedia hallucination.
(Score: 2) by c0lo on Thursday June 12, @08:27AM (3 children)
Irrelevant from the perspective of squeezing all the experimental data/info that can be achieved in the bounds of a technology available at a given time. ITER is not meant to produce a commercial fusor as such, it's a research and engineering megaproject [wikipedia.org], meant as a platform to test various materials, technologies and techniques for TOKAMAK fusors.
To date, none of the fusion projects has a winner for commercial energy production, no matter the approach to confinement/ignition.
Using AI in the feedback loop has high potential to lead to a commercial fusor [nature.com], it's good to have a platform to test/validate it and learn some more [iter.org]. Iff it turns out that the commercial fusion power requires AI (everything else failed), then one can make an argument that the transistor needed 70y of development to reach the stage in which it can be used for AI (nanoscale, high density, low power - otherwise AI could not reach enough functionality to control fusion). I'm not making this argument, it's just another speculation but, for now, so are your arguments.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @11:57AM (2 children)
And not irrelevant from the point of view of doing something useful with that technology! What makes your point of view better?
How many are trying? Glancing through the Wikipedia list [wikipedia.org] of past and present fusion power projects/experiments, I only see three companies in that list: entries from UK, China, and US (the last being a research project which has gone on since 1986). There's at least one new US entry building such, but it's not active yet.
My point here is that a commercial power plant will require an obsessive focus on the economics of making it work, not endless experimental data on systems that are too expensive to be made commercial.
(Score: 2) by c0lo on Thursday June 12, @01:50PM (1 child)
Many. So oppressive were those "institutional obstructions" that only US has at least 5 projects aiming towards commercial fusion power:
Helion Energy [wikipedia.org] - magneto-inertial confinement - mainly VC funding in $1e8-1e9 range [wikipedia.org] Sam Altman
Commonwealth Fusion Systems [wikipedia.org] managing SPARC [wikipedia.org] - tokamak - VC, Bill Gates, MIT in the picture, some Italian investors too
Lawrenceville Plasma Physics [lppfusion.com] - Dense Plasma Focus [wikipedia.org] - patents, private funding [tracxn.com]
TAE Technologies [wikipedia.org] - field reversed configuration/spheromak - started in stealth mode from 1998 to 2015, many patents and zillions publications in sci journals since, private capital, Google/Alphabet Inc
Zap Energy [wikipedia.org] - stabilized Z-pinch [wikipedia.org] - private held, with research contribution from University of Washington and Lawrence Livermore National Laboratory, many VC funding rounds in the $100M range
Oh, you and your obsessions. Cool down, mate, the world is large enough for both approaches.
https://www.youtube.com/@ProfSteveKeen https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday June 12, @07:19PM
Fusion power plants are the killer app for fusion (well once you get past thermonuclear bombs, that is).
(Score: 2, Insightful) by PiMuNu on Wednesday June 11, @11:08AM
> While the U.S. government spent approximately $800 million a year on fusion efforts during the Biden administration,
> China is investing more than twice that annually, IEEE Spectrum and others report.
It's worth adding at least a factor 2 due to Chinese employment & safety regulations, which are much less stringent than equivalent Western systems.
(Score: 5, Insightful) by gnuman on Wednesday June 11, @11:09AM
I guess that's the pitch. Looking for $$$ and throwing scary China around is their attempt. Maybe it will stick, maybe not.
Reality is that China can run their own *science* projects. And if they manage to create something, it will be their achievement because they did the science. In America, well, it seems the priority is to cut science funding. So no, you will probably not get your money especially since priority for Trump seems to be:
1. anti-science -- who needs that anyway?
2. pro-coal digging -- Biden was against, so we must *clean* that coal!
3. anti-renewables -- Biden was pro, so have to be against here!
(Score: 3, Funny) by Anonymous Coward on Wednesday June 11, @02:23PM
We need to be investing in coal mining technomology, like better spades and less health regulations. That's what will make us GREAT again! Fusion, superconductors, lol bullshit. Let's drill that shit 101!!
(Score: 1, Interesting) by Anonymous Coward on Wednesday June 11, @02:30PM
We need "Idiocracy" moderation
(Score: 2) by VLM on Wednesday June 11, @03:19PM (8 children)
China is good at spending money to "catch up" see the empty abandoned city stories.
Note how incredibly careful they are to avoid discussing any progress or results, merely spending figures.
My guess is they've caught up to 1990, maybe Y2K. It must be very embarrassing if they're censoring it this thoroughly.
(Score: 2, Touché) by VLM on Wednesday June 11, @03:23PM (1 child)
Oh I see our welfare recipients want more money to accomplish nothing other than the reception of money. That's why they were so careful to avoid discussing results. The whole story makes sense now.
I'm sure WRT spending figures you can trust everything that comes out of China, just like everything else that comes out of China, LOL.
(Score: 3, Insightful) by gnuman on Wednesday June 11, @09:32PM
It's not from China. It's from outside of China about China...
Ben Levitt, the head of R&D for Zap Energy, speaking on a fusion panel at the Seattle Investor Summit+Showcase.
Brian Riordan, co-founder and chief operating officer of Avalanche Energy
Anthony Pancotti, co-founder and head of R&D for Helion Energy.
It's the same when https://en.wikipedia.org/wiki/Missile_gap [wikipedia.org] -- it's FUD to bamboozle you.
Like I wrote, China has a right to run their own science experiments. They have enough know-how to do so and have rights to pursue their own tech. Heck, China and Russia are members of https://en.wikipedia.org/wiki/ITER [wikipedia.org] and even US is there. So, it seems this is even more FUD from the private sector. Reality is that fusion is hard, and it's not even the fusion part that makes it hard -- it's all the other stuff around fusion, like vessel longevity.
(Score: 5, Interesting) by zocalo on Wednesday June 11, @04:17PM (4 children)
Don't get me wrong, there's totally attempted censorship of project failures and other things that underperform and don't go well, and I don't doubt the figures on some of the favourable results are inflated too, but that propaganda sword cuts both ways; the west is also downplaying just how close, or even ahead, in many areas China actually is, presumably because that would look bad when set against the "We're Number #1" rhetoric. Sure, some of that tech and innovation was stolen, but they're now well out of the "developing economy" phase and more than capable of doing their own innovation when needed.
China's trajectory has been pretty clear for some time now; burying our collective heads in the sand and chanting "we're number #1" isn't going to change that, and certainly won't have a happy ending. We (that is every developed country outside China) need to be building on our strengths and pouring money into innovation in fields that are on the cutting edge and will have a clear global market in the future. The US can dump energy efficiency standards, environmental regulation, and "drill, baby, drill" all it likes, but the rest of us are most likely going be buying a lot of clean, green, tech in the future and there's no way anyone will be buying Trump's US-made products in that market if they don't even have have any to sell.
UNIX? They're not even circumcised! Savages!
(Score: 3, Informative) by corey on Wednesday June 11, @09:06PM (3 children)
You’re right. I’ve been to China and seen the ingenuity and scale they do things. They can go from CAD file to physical prototype of a PCB with fitted electronics in a day, at a facility in the same city. The test it, tweak it, send it back to the manufacturer on day three for another 8hr turnaround. They need this in California or Texas. Yeah quality might be an issue but it’s getting better. Like Japan decades ago.
I think this story is good. I’m glad to hear the world superpowers competing for something good for the world (unlike AI). Let them optimise fusion and scale it, we need the power.
(Score: 2) by VLM on Friday June 13, @03:14PM (2 children)
Unironically, you can do that today in Tualatin Oregon. The price is higher but it works. I am vaguely confused if PCB-Unlimited is in Tualatin or thats just some mailing address and the factory is somewhere else. If you demand weird parts there might be supply chain disclaimer issues. No different than ordering from China where the odds of the parts not being counterfeit are ... low.
You probably won't like the price that much. People talk about how they want this and want that but if they won't pay for it, it doesn't really matter if someone offers the product or not.
Another issue that comes up is there's no need for middlemen in the USA; if you REALLY need rapid turn around prototyping, you set up a micro-fab in the lab and make your own PCBs, the US has the capital resources for a company to simply buy the equipment rather than insist on paying someone else a profit to buy the equipment for you. This also works really well with ultra-deep pockets like military and aerospace, no classified chain of custody issues if the development board was made in house.
(Score: 2) by corey on Sunday June 22, @11:13AM (1 child)
Initially I was going to say PCB Unlimited were a shop front for Chinese fabrication, as a bunch of places are here in Australia. But I checked them out, they claim to have a USA made fab as well as China and Taiwan. So that’s pretty cool — their capability is pretty good too, advanced process up to 40 layers, they do Rogers etc too. And they probably do half decent upfront DFM too, unlike PCBWay etc.
https://www.pcbunlimited.com/products/usa-quickturn-pcbs [pcbunlimited.com]
(Score: 2) by VLM on Sunday June 22, @03:13PM
They can turn a zip file into physical hardware faster than a one-way subsonic aircraft, as long as you stick to reasonable jellybean parts (not some weird stuff).
What kills the business model of places like this is any department that can afford a couple EE salaries in the USA can easily afford the capex to set up a small lab and hire a tech to baby sit it. So if national security depends on going from CAD model to hardware in two hours, you can do it for the cost of a really nice new car (like a van or truck, maybe). And if you can wait "a couple days" the cost of having China do it is a rounding error compared to salaries. And there's just not much in between, but there are companies doing it.
There's an issue of mass production of the machinery. Theres not many sales for small pick and place machines so you can buy a steel cutting milling machine from Haas for 2/3, maybe 1/2 the price of a similar level of professionalism pick and place. Weird situation in the market that turning big chunks of titanium into chips to ten thousandths of an inch accuracy is cheaper than dropping a tiny chip capacitor into place accurate to only a thousandth of an inch.
(Score: 2) by PiMuNu on Thursday June 12, @09:32AM
> My guess is they've caught up to 1990, maybe Y2K. It must be very embarrassing if they're censoring it this thoroughly.
In my field they have built to about 2000 and are working on 2020 in the next 5 years.
Extrapolation is dangerous; but that puts dt/dt around 4
(Score: 5, Insightful) by Mojibake Tengu on Wednesday June 11, @06:30PM (3 children)
Contrary to the popular belief of bankers and finance jugglers, you can't build industry out of pure money. Any industry.
You need knowledge, skills, and that means you need to create a pool of skilled labourers imbued with that knowledge and skills.
Well, I am not impressed by current U.S. technology or education model either. Though it's their problem, not mine.
Rust programming language offends both my Intelligence and my Spirit.
(Score: 3, Touché) by Anonymous Coward on Wednesday June 11, @09:20PM
"Contrary to the popular belief of bankers and finance jugglers, you can't build industry out of pure money. "
But you can scam more money by just manipulating money.
They have demonstrated anything else, including any other industry, is uninteresting for the purpose of making money.
(Score: 3, Funny) by PiMuNu on Thursday June 12, @09:35AM
Anyone who believes in "FTE" (Full Time Equivalent) is an idiot*
*I use FTE estimates all the time. I just don't believe them.
(Score: 2) by VLM on Friday June 13, @03:23PM
That's another problem, the USA already creates far more laborers than there are jobs for them. My guess is the "problem" is China has 2 million engineering jobs and only 1.5 million grads, well, the USA has the flip side of that problem.
You can "create" a million BSCS grads but if there's only jobs for 250K of them, you can't fix that mismatch by demanding they "create" two million BSCS grads next year. All you'll get is 2 million grads and only 250K jobs for them. "Well maybe if we had 3M grads the problem would fix itself" uh no.
This is a big problem in EE land where for generations we've heard "there's not enough EE grad to compete" in the USA but most of them can't get EE jobs so they become programmers. I worked with a lot of sysadmin / security / IT / software dev / FPGA programmers / sales engineers with EE degrees who simply can't get EE jobs, there just aren't any WRT supply and demand.