Image: The Joint European Torus, a fusion machine in the U.K., holds the record for Q performance among tokamaks. This image shows its interior with a hot plasma overlaid. Credit: UKAEA, courtesy of EUROfusion
I’ve detailed three of the six milestones on the path to commercial fusion energy. Now it’s time for the fourth, a truly monumental achievement called net fusion energy. It’s the first real demonstration that you’ve created a system that can actually make the basis of a power plant — you’re not just on the path to fusion energy, but you’re actually making it at levels that are useful.
In scientific circles, this milestone is called Q>1. Q is a ratio showing the fusion performance in the superhot cloud of charged particles called a plasma. Specifically, Q is the energy that the fusion process creates by combining the plasma’s nuclei, divided by the energy needed to heat the plasma at those conditions. When Q exceeds 1, you’ve got a machine that creates net fusion energy — in other words, more energy out than in, at the heart of the machine.
Achieving Q>1 means you must overcome the most scientifically challenging parts of a fusion power plant. It proves your physics works and that you can make the fusion process overcome its losses. These conditions are very challenging, and thus these machines are not for the faint of heart. As a result, reaching Q>1 also shows you have the organization and the engineering capability to integrate many subsystems — power, fueling, heat management, diagnostics, and more — into a functioning fusion machine. That’s why it’s such a powerful step on the path to bringing fusion’s clean, essentially limitless power to the electricity grid.
To put Q>1 into context, let’s catch up on the first three milestones toward putting competitive fusion energy on the electricity grid:
- Creating a stable plasma, the hot, hard-to-control cloud of electrically charged particles that’s the fuel for a fusion machine
- Heating that plasma to 10 million degrees Celsius, hot enough to see the complications a plasma throws at you
- Scoring high on a measurement called the triple product that shows you can keep your plasma hot enough, dense enough, for long enough to approach fusion conditions
If you keep increasing your triple product, you’ll reach conditions where, if you use the right fuels, you’ll get Q>1. This isn’t the end of the journey. But once you’ve reached this step, it’s relatively easy to improve performance so your plasma reaches higher levels, and that lets you start to work on harnessing that fusion energy to put it to work selling it outside the plant itself.
What is Q>1 and how does it relate to the power plant?
The hard part about fusion is the plasma. It’s the state of matter we as humans know the least about, and getting it to the conditions for the fusion reaction is very difficult. It needs to be hot, dense, and insulated enough to show a good triple product. This isn’t to say that the rest of the engineering challenges are easy. It’s just that comparatively speaking, humans are earlier in their journey on the plasma itself: We’re still discovering things and figuring out how to best harness it. That’s why we have plasma physics as a scientific discipline.
Q is still all about the plasma, but it’s no longer just about attaining the conditions needed for fusion reactions to occur. Instead, it’s about the actual fusion reactions. Q is the ratio of energy produced by the fusion — something that’s relatively easy to understand — divided by the energy used to keep the plasma hot. That part is a bit more nuanced. I’ll break it down here.
First, measuring the fusion energy created by the plasma means you’re using real fusion fuels to make that fusion energy. This isn’t about the triple product anymore.1 Second, you can think of the plasma in a fusion machine as its own thing — a blob in a vacuum that’s so well insulated from the rest of the world that it’s easy to see where it starts and ends.2 Because of that isolation, you can easily measure the energy you send into the plasma to heat it to fusion conditions. At Q>1 we’ve now graduated from pure plasma physics to the combination of plasma physics and nuclear physics.3
Drawing a boundary to measure Q
Sometimes the energy used to heat the plasma gets confused with the energy used to run the systems that heat the plasma or the energy used to run the plant. But Q is just about the plasma itself, not about these systems. You have to draw the boundary somewhere, and the logical location is at the plasma edge because plasmas are the hardest, most novel, and most easily separable thing in the system.
Think of it like boiling a pot of water on the stove for dinner. You can calculate the amount of heat that went into the water. That’s less than the amount of heat the stovetop provided — some of it leaks out the edges into air — and you need other energy needed to keep the lights on to see the water boil. And all that is less than the amount of energy a power plant generates to send to the house. You need these later things to make dinner, but they’re far removed from heating the water itself.
In fusion, the energy that goes into the plasma comes from lasers or microwaves or radio waves or changing magnetic fields or firing particles into the plasma. To generate each of these, you have to engineer equipment that’s not perfectly efficient. Then the fusion reactions are converted to electricity, and that’s not perfectly efficient. And some of the electricity goes back to power the plasma heating, being recirculated. Thus, the amount of electricity left to sell depends not just on the Q, but on all these engineering efficiencies and choices of equipment to heat the plasma and convert the energy from the plasma. The people who engineer those systems are very different from the people who handle the plasmas. The understanding levels are very different, and the options are very different. Thus we separate all the engineering systems supporting the plant and focus on the plasma itself. A reasonable Q isn’t sufficient to get to a power plant, but it is necessary, and it is separable.4
This isn’t to say that you can’t make net electricity to sell at low Q. Indeed, if your supporting subsystems are very, very efficient you can have a little bit extra to sell at low Q. But at high Q, you can have a lot of extra to sell for the same capital intensive plant you built. That’s why almost all fusion power plant developers aim for Q>1 and most aim for Q>10.5
Attaining Q>1 isn’t a panacea, but it is an inflection point. And, because of the nature of the nuclear reactions, once you’re above Q>1, the system gets easier and easier to run as the plasma starts to heat itself from the fusion reactions.6 You can see this in the triple product chart: The lines of Q get closer and closer together as you get higher and higher. The distance between Q=0.1 and Q=1 is much larger than the distance between Q=1 and Q=10, and that distance is much larger than Q=10 and Q=100. Once you’re above Q>5, your machine will be dominated by the fusion reactions themselves and will be on the downhill slope to higher and higher performance.
The fourth milestone uses Q because it indicates a foundational level of fusion capability. Primarily, it’s the core achievement that shows your physics approach, your machine, and your company have successfully built the foundation for a fusion power plant. But also, it’s widely accepted and understood in the fusion research community, measured by many fusion projects, published in many scientific papers, and spotlighted as a milestone in reports like the National Academy of Sciences’ Bringing Fusion to the U.S. Grid.
Importantly, beyond the plasma itself, by making a plasma that’s Q>1 you’re in a serious regime, making a bunch of fusion power. It might not be enough to sell electricity or make money, but the plasma is starting to heat itself up and the power flows have a bunch of fusion energy in them where before they had little. You’re using fusion fuel, you’re making lots of energy, you’re dealing with neutrons and radiation.
The thing is starting to look more and more like a power plant where before it looked like a physics experiment.
NIF, the only Q>1 machine so far
Several types of fusion machines have made it past these first three milestones. But only one machine in the world has shown Q>1: the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory. It crossed this threshold for the first time on Dec. 5, 2022 with Q of 1.5. They’ve reached even higher since then including Q of 4.13 in April 2025.
When NIF surpassed Q>1, the world took notice. It was the culmination of decades of work to build the facility, overcome problems, and improve performance. The achievement validated our understanding of the core physics of fusion and brought new attention to the commercial fusion energy industry. Plasma physics is plasma physics, and NIF showed we do indeed understand it well enough to get to the conditions needed for plasma to be the basis of a fusion power plant.
NIF is a research facility designed to study nuclear weapons, though, not a part of an effort to put power on the grid. Its engineering systems aren’t optimized to get the most out of the fusion power created, or to make a lot of fusion power, or to be efficient in heating the plasmas. That’s okay — it’s not why NIF was built. Now people are working on those systems with the knowledge that the plasma itself is sound. There are of course other types of fusion systems that are optimized from the beginning to be fusion power plants, with subsystems that are much more efficient and with power plant scaling built into the concept from the beginning.
When a commercial fusion company passes Q>1, it’ll be another moment to celebrate. Here at Commonwealth Fusion Systems (CFS), we aim to reach Q>1 with our SPARC machine in about two years. We expect that’ll be the first time a commercially relevant machine crosses this threshold — a major moment of proof that fusion energy can meet its world-changing potential.
Progress on the path to Q>1
Fusion companies are developing several different types of fusion devices to enable fusion and generate power. NIF aims 192 powerful lasers at a pea-sized pellet of fuel. Our device type, called a tokamak, uses strong magnets to confine and control the plasma. Other ideas include stellarators, field-reversed configurations, Z-pinch machines, and magnetic mirrors.
The laser compression approach is the only one that’s made it to the Q>1 threshold. The next closest are the tokamaks. The closest any tokamak has come to net fusion energy was the Joint European Torus (JET) in the UK with Q of 0.63 in 1997. Everything else is much farther behind in this parameter.
If you want to see how well fusion machines have fared over the years, I recommend the 2025 research paper by Sam Wurzel and Scott Hsu that tracks progress. Look for the column labeled Qsci in Table IV. Their preceding paper from 2022 also details many of the factors involved.7
How to measure net fusion energy
To calculate Q, it’s crucial to accurately measure both the power the machine uses to heat the plasma and the fusion power it produces. That requires diagnostic equipment that’s well established and well operated, just as with the previous three fusion milestones. Generally, there’s a careful accounting of all the energy flowing into the plasma — for example, the energy delivered in laser light, the energy coupled from microwaves, or the energy sent in from the magnets. And because this all crosses a vacuum chamber before it hits the plasma, we usually add it up at that boundary. And then measuring the power from the fusion reactions is done using carefully calibrated neutron diagnostics.
SPARC will use multiple methods proven in earlier tokamaks and validated in peer-reviewed research papers. For example, because fusion produces neutrons, we’ll have neutron-counting equipment that offers a direct view of how much fusion is occurring.
As for measuring the power we’re pumping into SPARC to sustain the fusion process, we have two systems to track. We’ll monitor the electrical current we drive into SPARC’s central magnet, which creates another current inside the plasma and heats the plasma resistively, and we’ll monitor the radio power we pipe into the plasma.
Moving beyond Q>1 toward a fusion power plant
Achieving Q>1 is a foundation you build upon to bring fusion energy to the grid, the job of the plasma is nearly done. For the next milestone, you have to go beyond that — running all the supporting equipment at high enough efficiency, capturing all that fusion power and turning it to electricity, and having power-plant like plant support systems. That’s what milestone five is about: net electricity production. In other words, having something extra to sell. It requires a power plant — our ARC design, for example — not just a demonstration machine like SPARC. Stay tuned for details on that next step soon.
Footnotes
- Sometimes you see the term Qequivalent. This means that the plasma is at conditions where if you were using real fusion fuels like deuterium and tritium, your machine would make enough fusion power to satisfy Q>1 — but you’re not actually using real fusion fuels. So Qequivalent>1 is just about the triple product, not about fusion energy production. Take this measurement with a grain of salt, and no, it doesn’t count in this milestone. ↩︎
- If the plasma touches anything, you’re very likely back at the second milestone, because the plasma is leaking out its power and cooling down. ↩︎
- You can base Q either on power or on energy. You can measure Q in a pulsed system where you integrate all the fusion energy created in the pulse, then integrate all energy used to create and heat the plasma, then divide them. That’s how NIF and magneto-inertial fusion systems calculate Q. You can also measure Q as a steady-state system based on the power coming from the fusion reactions divided by the power used to keep the plasma hot. That’s what stellarators and tokamaks do. In the limit of long pulses, the energy used to create the plasma is small compared to the energy used to keep it hot over the length of the pulse, which means the two approaches reduce to the same Q value. There are subtleties, as in all things in fusion, but conceptually it’s not hard to see the basics. ↩︎
- I’ll get into all the math in the next milestone. But here’s a thought experiment. Suppose you had a laser that’s 1% efficient. In that case, a unit of energy delivered to the plasma would need 100 units of energy to power the laser heating the plasma. At Q>1, you’d have nothing to sell. But if the laser is 100% efficient, then for every unit of energy delivered to the plasma it only took a single unit of energy to power the laser. At Q>1 you’re close to having something extra to sell. The laser changed, but in both cases, the plasma did the same thing. Doing laser engineering is separate from doing plasma physics; you should go work on the laser next because the plasma is in the right regime. ↩︎
- At Q=1, you get two units of energy out of the plasma for every unit you put in: one unit from the fusion power, and one unit from the heat you put in; energy isn’t destroyed. And at Q=10 you get 11 units out for every unit in: 10 from the fusion power and the one you put in. So you can see how higher Q makes life easier for the plant designer. ↩︎
- At Q=1, all energy from fusion reactions totals the energy used to heat the plasma. Sometimes, you’ll see Q=5 as an important threshold called a “burning plasma” because in deuterium-tritium (DT) fusion, 20% of the energy comes out as charged particles that stay in the plasma, helping to heat it. Thus at Q=5, the energy heating the plasma from the fusion reactions themselves is equal to the energy provided to heat the plasma from external sources. The Q for a burning plasma depends on the fuel you use because each fuel has a different split of energy into charged and neutral particles. Another important threshold is Q=infinity, called ignition. This condition means that plasma is heating itself and you don’t have to provide any external heating. The sun is ignited: It keeps itself going by burning its fuel without any extra energy source. So is a flame in a candle: You can remove the lighter and it will keep burning on its own fuel without help. ↩︎
- When looking through information like this, you’ll sometimes see other terms for Q like energy gain and net gain. They’re all roughly the same. Sometimes you’ll see Qeng — that’s about the ratio of energy from fusion to the ratio of energy to run the whole plant. We avoid this term because it’s confusing and not terribly useful for thinking about a power plant. There, the important thing is the absolute magnitude of the energy you have left to sell, and each system has different scalings and risks. More on this in the next milestone. ↩︎
