This is a great update! I hope the authors continue publishing new versions of their plots as the community builds up towards facility gain. It's hard to keep track of all the experiments going on around the world, and normalizing all the results into the same plot space (even wrt. just triple product / Lawson criteria) is actually tricky for various reasons and takes dedicated time.
Somewhat relevant, folks here might also be interested in a whitepaper we recently put up on arXiv that describes what we are doing at Pacific Fusion: https://arxiv.org/abs/2504.10680
Section 1 in particular gives some extra high-level context that might be useful to have while reading Sam and Scott's update, and the rest of the paper should also be a good introduction to the various subsystems that make up a high-yield fusion demonstration system (albeit focused on pulser-driven inertial fusion).
I heard that NIF was never intended to be a power plant, not even a prototype of one. It's primarily a nuclear weapon research program. For a power plant you would need much more efficient lasers, you would need a much larger gain in the capsules, you would need lasers that can do many shots per second, some automated reloading system for the capsules, and you would need a heat to electricity conversion system around the fusion spot (which will have an efficiency of ~1/3 or so).
It's an experimental facility. Yes, a power plant would need much more efficient lasers, but NIF's lasers date back to the 1990s, equivalent modern lasers are about 40X more efficient, and for an experiment it's easy enough to do a multiplication to see what the net result would have been with modern lasers.
Modern lasers can also repeat shots much more quickly. Power gain on the capsules appears to scale faster than linear with the input power, so getting to practical gain might not be as far off as it appears at first glance.
These are some of the reasons that various fusion startups are pursuing laser fusion for power plants.
From what I understood, laser fusion needs laser efficiencies not just 40x better than what NIF uses, but like 3 or 4 orders of magnitude more efficient than the state of the art. Seems like a non-starter.
I was trying to work out a joke about buying better lasers off of alibaba but it seems that despite being 30 years old they're still orders of magnitude beyond off the shelf options.
From my time in fusion research circles, you're correct, but it's also not a simple "weapons or energy?" question. It could only have ever been a pure research facility. At the time of design, the physics wasn't certain enough to aim for net energy gain. Where the weapons research came in is in the choice of laser focus. Instead of "direct drive", where the lasers directly strike the fusion fuel, NIF lasers strike the inside of a Hohlraum, which produces X-Rays that then heat the fuel. X-Ray opacity is an important topic in nuclear weapons research.
Bear in mind that I wasn't directly involved, and this my impression picked up from conversations during my time in fusion research, which was about 10 years ago.
>NIF is a key element of the National Nuclear Security Administration’s science-based Stockpile Stewardship Program to maintain the reliability, security, and safety of the U.S. nuclear deterrent without full-scale testing.
Yes, after the test ban treaties, there was a huge push into exploring mathematical emulations of all aspects of fusion, and all assorted bombs, as well as laser ignition of pellets with these large lasers using inertial confinement of the pellet as the laser impacted it - analysing the fusion by observation of emitted neutrons. xrays etc. They issued reports from time to time(sanitised), and probably used the secret data to fine tune emulated weapons with fact points. The pellets were composed of potential fuels, various Hydrogens and Lithiums, varied in composition to explore the ignition space. A number of pellets performed well in terms of gain, but were far-far from useable fusion when the LL labs costs were factored in. I think they determined it could not ever work as a fusion energy source, but it provided data. They still mine data from it with various elemental mixes making up the pellets.
The primary purpose of the NIF is to maintain the US nuclear stockpile without nuclear tests. The lasers very inefficient (iirc about 2%). The success they claimed is that the energy released by the burning plasma exceeds the laser energy put into the fuel capsule. Since NIF was never intended to be a power plant they don't use the most efficient lasers.
Nothing about the NIF looks like a power plant to me. It's like the laser weapons guy and the nuclear weapons guy found a way to spend giant piles of money without having to acknowledge the weapons angle.
A lot of people think so, but the US government openly spends way more money on nuclear weapons than on fusion research. We'll spend almost a trillion dollars on nuclear weapons over the next decade.[1] The government's fusion funding was only $1.4 billion for 2023.[2]
So it seems more likely to me that some physicists figured out how to get their fusion power research funded under the guise of weapons research, since that's where the money is. NIF's original intent was mostly weapons research but it's turned out to be really useful for both, and these days, various companies are attempting to commercialize the technology for power plants.[3]
Yes. The NIF is a weapons research lab, not a power research lab.
The purpose of it is to show that the USA is still capable of producing advanced hydrogen bombs. More advanced then anybody else.
The '2.05 megajoules' is only a estimation of the laser energy actually used to trigger the reaction. It ignores how much power it took to actually run the lasers or reactor. Even if they update the lasers with modern ones there is zero chance of it ever actually breaking even. It is a technological dead end as far as power generation goes.
The point of the 'breakthrough' is really more about ensuring continued Congressional approval for funding then anything else. They are being paid to impress and certainly they succeeded in that.
However I suspect this is true of almost all 'fusion breakthroughs'. They publish updates to ensure continued funding from their respective governments.
People will argue that this is a good thing since it helps ensure that scientists continue to be employed and publishing research papers. That sentiment is likely true in that it does help keep people employed, but if your goal is to have a working and economically viable fusion power plant within your lifetime it isn't a good way to go about things.
If the governments actually cared about CO2 and man-made global warming they would be investing in fusion technology and helping to develop ways to recycle nuclear waste usefully. Got to walk before you can run.
It was never intended to be a power plant but it was hoped that it would achieve a net gain fusion reaction for the first time. This turned out to be a lot harder than expected.
It should be noted that "breakeven" is often misleading.
There's "breakeven" as in "the reaction produces more energy than put into it", and there's breakeven as in "the entire reactor system produces more energy than put into it", which isn't quite the same thing.
We are careful to always specify what kind of “breakeven” or “gain” is being referred to on all graphs and statements about the performance of specific experiments in this paper.
Energy gain (in the general sense) is the ratio of fusion energy released to the incoming heating energy crossing some closed boundary.
The right question to ask is then: “what is the closed boundary across which the heating energy is being measured?”
For scientific gain, this boundary is the vacuum vessel wall. For facility gain, it is the facility boundary.
It's always confused me a bit. It's not like if you put 10kWh into the reactor, that 10kWh goes away. You still lose a significant fraction of it in inefficiency of the cycle but it still goes towards heat which can be used to heat steam and turn a turbine. iirc, you can get about 4kWh back.
On the other side of the coin, if you put 10kWh in and get 10kWh of fusion out, that's 20kWh to run a steam turbine, which nets you about 8kWh. So really you need to be producing 15kWh of heat from fusion for every 10kWh you put in to break even.
Why is the last plot basically empty between 2000 and 2020? I understand that NIF was probably being built during that time, but were there no significant tokamak experiments in that time?
Author here - some other posters have touched on the reasons. Much of the focus on high performing tokamaks shifted to ITER in recent decades, though this is now changing as fusion companies are utilizing new enabling technologies like high-temperature superconductors.
Additionally the final plot of scientific gain (Qsci) vs time effectively requires the use of deuterium-tritium fuel to generate the amounts of fusion energy needed for an appreciable level of Qsci. The number of tokamak experiments utilizing deuterium tritium is small.
Thanks a lot for this research. Seing the comments here I think it's really important to make breakthroughs and progress more visible to the public. Otherwise the impression that "we're always 50 years away" stays strong.
Mind you, it's not useless! It produced a TON of very useful fusion research: neutral beam injectors, divertors, construction techniques for complex vacuum chambers, etc. At this point, I don't think it's going to be complete by the time its competitors arrive.
One spinoff of this is high-temperature superconductor research that is now close to producing actually usable high-TC flexible tapes. This might make it possible to have cheaper MRI and NMR machines, and probably a lot of other innovations.
> actually usable high-TC flexible tapes. This might make it possible to have cheaper MRI and NMR machines, and probably a lot of other innovations.
I'm sure there'll be plenty of fascinating applications of high-Tc tape, however I'm not sure MRI/NMR machines will be one of those. There would still be a lot of thermal noise due to the high temperature. Which is why MRI/NMR machines tend to use liquid helium cooling, not because superconductors capable of operating at higher temperatures don't exist.
ITER doesn't use high temperature superconductors. It uses niobium-tin and niobium-titanium low temperature superconductors in its magnets.
ITER has been criticized since early days as a dead end, for example because of its enormous size relative to the power produced. A commercial follow-on would not be much better by that power density metric, certainly far worse than a fission reactor.
There is basically no chance than a fusion reactor operating in a regime similar to ITER could ever become an economical energy source. And this has been known since the beginning.
I call things like ITER "Blazing Saddles" projects. "We have to protect our phony baloney jobs, gentlemen!"
Presumably because everyone in MCF has been waiting for ITER for decades, and JET is being decommissioned after a last gasp. Every other tokamak is considerably smaller (or similar size like DIII-D or JT-60SA).
Much of the interesting tokamak engineering ideas were on small (so low-power) machines or just concepts using high-temperature superconducting magnets.
Anyone have any idea where First Light Fusion's third machine fits into this?
The idea of using literal guns (gunpowder, then light gas gun, then coil gun) to impact projectiles against each other seemed like it was probably ludicrous, but I haven't seen any critical media or numbers yet.
This will probably need to be updated soon. There are rumors NIF recently achieved a gain of ~4.4 and ~10% fuel burn up. Being able to ignite more fuel is notable in and of itself.
In the context implied above it is the ratio of fusion energy released to laser energy on target or the laser energy crossing the vacuum vessel boundary (they are the same in this case). So it would have been more precise to say "target gain" or "scientific gain".
Readit News
Somewhat relevant, folks here might also be interested in a whitepaper we recently put up on arXiv that describes what we are doing at Pacific Fusion: https://arxiv.org/abs/2504.10680
Section 1 in particular gives some extra high-level context that might be useful to have while reading Sam and Scott's update, and the rest of the paper should also be a good introduction to the various subsystems that make up a high-yield fusion demonstration system (albeit focused on pulser-driven inertial fusion).
Any truth to that?
Modern lasers can also repeat shots much more quickly. Power gain on the capsules appears to scale faster than linear with the input power, so getting to practical gain might not be as far off as it appears at first glance.
These are some of the reasons that various fusion startups are pursuing laser fusion for power plants.
Dead Comment
Bear in mind that I wasn't directly involved, and this my impression picked up from conversations during my time in fusion research, which was about 10 years ago.
https://lasers.llnl.gov/about/what-is-nif
>NIF is a key element of the National Nuclear Security Administration’s science-based Stockpile Stewardship Program to maintain the reliability, security, and safety of the U.S. nuclear deterrent without full-scale testing.
So it seems more likely to me that some physicists figured out how to get their fusion power research funded under the guise of weapons research, since that's where the money is. NIF's original intent was mostly weapons research but it's turned out to be really useful for both, and these days, various companies are attempting to commercialize the technology for power plants.[3]
[1] https://theaviationist.com/2025/04/26/us-nuclear-weapons-wil...
[2] https://www.fusionindustryassociation.org/congress-provides-...
[3] NYTimes: https://archive.is/BCsf5
The purpose of it is to show that the USA is still capable of producing advanced hydrogen bombs. More advanced then anybody else.
The '2.05 megajoules' is only a estimation of the laser energy actually used to trigger the reaction. It ignores how much power it took to actually run the lasers or reactor. Even if they update the lasers with modern ones there is zero chance of it ever actually breaking even. It is a technological dead end as far as power generation goes.
The point of the 'breakthrough' is really more about ensuring continued Congressional approval for funding then anything else. They are being paid to impress and certainly they succeeded in that.
However I suspect this is true of almost all 'fusion breakthroughs'. They publish updates to ensure continued funding from their respective governments.
People will argue that this is a good thing since it helps ensure that scientists continue to be employed and publishing research papers. That sentiment is likely true in that it does help keep people employed, but if your goal is to have a working and economically viable fusion power plant within your lifetime it isn't a good way to go about things.
If the governments actually cared about CO2 and man-made global warming they would be investing in fusion technology and helping to develop ways to recycle nuclear waste usefully. Got to walk before you can run.
There's "breakeven" as in "the reaction produces more energy than put into it", and there's breakeven as in "the entire reactor system produces more energy than put into it", which isn't quite the same thing.
Energy gain (in the general sense) is the ratio of fusion energy released to the incoming heating energy crossing some closed boundary.
The right question to ask is then: “what is the closed boundary across which the heating energy is being measured?” For scientific gain, this boundary is the vacuum vessel wall. For facility gain, it is the facility boundary.
On the other side of the coin, if you put 10kWh in and get 10kWh of fusion out, that's 20kWh to run a steam turbine, which nets you about 8kWh. So really you need to be producing 15kWh of heat from fusion for every 10kWh you put in to break even.
Availability (reliability engineering) https://en.wikipedia.org/wiki/Availability
Terms from other types of work: kilowatt/hour (kWh), Weight per rep, number of reps, Total Time Under Tension
Additionally the final plot of scientific gain (Qsci) vs time effectively requires the use of deuterium-tritium fuel to generate the amounts of fusion energy needed for an appreciable level of Qsci. The number of tokamak experiments utilizing deuterium tritium is small.
Here was my completely layman attempt to forecast fusion viability a few months ago. https://news.ycombinator.com/item?id=42791997 (in short: 2037)
Is there some semblance of realism there you think?
Mind you, it's not useless! It produced a TON of very useful fusion research: neutral beam injectors, divertors, construction techniques for complex vacuum chambers, etc. At this point, I don't think it's going to be complete by the time its competitors arrive.
One spinoff of this is high-temperature superconductor research that is now close to producing actually usable high-TC flexible tapes. This might make it possible to have cheaper MRI and NMR machines, and probably a lot of other innovations.
I'm sure there'll be plenty of fascinating applications of high-Tc tape, however I'm not sure MRI/NMR machines will be one of those. There would still be a lot of thermal noise due to the high temperature. Which is why MRI/NMR machines tend to use liquid helium cooling, not because superconductors capable of operating at higher temperatures don't exist.
ITER has been criticized since early days as a dead end, for example because of its enormous size relative to the power produced. A commercial follow-on would not be much better by that power density metric, certainly far worse than a fission reactor.
There is basically no chance than a fusion reactor operating in a regime similar to ITER could ever become an economical energy source. And this has been known since the beginning.
I call things like ITER "Blazing Saddles" projects. "We have to protect our phony baloney jobs, gentlemen!"
Much of the interesting tokamak engineering ideas were on small (so low-power) machines or just concepts using high-temperature superconducting magnets.
There's the common joke that fusion is always 30 years away, but now with the help of ITER, it's always 10 years away instead.
The idea of using literal guns (gunpowder, then light gas gun, then coil gun) to impact projectiles against each other seemed like it was probably ludicrous, but I haven't seen any critical media or numbers yet.
Deleted Comment
(it's been 30 years away for 50 years already, but as long as I'm not dead 30 years from now, it's still a good investment...)