Launch HN: Maritime Fusion (YC W25) – Fusion Reactors for Ships
Yes, we know: fusion has been the energy source of the future…and it always will be. But high-temperature superconductors (HTS) have changed the game for magnetic confinement, and we believe we’ll witness Q > 1 within a few (say 3) years. That’s huge.
(Side note: Q is the ratio of input power divided by output power. Q> 1 means the reactor is producing more power than it consumes, achieving ‘breakeven.’)
However, getting to breakeven is just the first daunting challenge. Making the first-of-a-kind (FOAK) reactors cost-competitive on the grid? That might be even harder than achieving breakeven.
That’s why we’re taking this soon-to-be breakthrough in fusion and applying it to the first market we believe makes sense: ships.
Instead of targeting 24/7 baseload grid electricity—where fusion has to compete with solar, wind, batteries, and natural gas—we’re focusing on large commercial shipping (>10,000 TEU) and mobile military vessels to provide ship-to-shore power capability.
Why ships? They don’t have great alternatives—the shipping industry is desperate to decarbonize. Hydrogen and ammonia are being explored, but come with serious downsides: low energy density, flammability, leaks, and massive infrastructure challenges. Fusion will provide a high-energy-density, long-range solution without the same infrastructure challenges—once it works, of course!
One common question is, why not fission? Fission works technically, but not practically. Small Modular Reactors (SMRs) could power ships, but licensing fission reactors on land is already brutally hard and expensive—doing it for vessels moving between international ports with enriched uranium is nearly impossible. Public perception is another major barrier: if we’re deploying thousands of nuclear reactors globally, they need to be meltdown-proof. Fusion is the only way to guarantee that. Regulation also isn’t as bad. While fusion won’t be a walk in the park to license, the NRC has declared a distinct framework for it—more like particle accelerators and hospitals than nuclear power plants. That’s a game-changer.
Instead of a 500+ MW grid-scale reactor, our system is 25 MWe, designed for ship propulsion. Our tokamak is roughly JET-sized, but with HTS magnets (8-9T) and higher plasma current (~10MA). The first-wall power flux is down from multi-MW/m² to nearly 500 kW/m²—still tough, but not nightmare mode. The materials challenges associated with the first wall and nuclear activation of the structures is greatly reduced. Also, ships don’t require 90% uptime like grid power plants. Downtime for maintenance is part of normal operations, making this a far more forgiving early application of fusion, unlike the grid where every down hour is lost revenue.
Jason and I come from SpaceX and Tesla, where we solved hard engineering problems at scale. My background is nuclear engineering (NC State, BS) and plasma physics (Columbia University, MS). We’ve been busy during our time in YC making technical progress on our reactor design, and are in the process of assembling a team of engineers who can pull this off.
This is a ridiculously hard problem, of course. But we think it’s the right hard problem—one that’s actually solvable (and worth solving!) with today’s tech if applied correctly. Eventually the cheaper and more robust SOAK and NOAK (second-of-a-kind and nth-of-a-kind) reactors will arrive in the coming decades (2050-2060) and then we'll pivot to decarbonising the grid and saving the world (we'll need to change our name), but until then we'll be out in the ocean!
Would love to hear your thoughts—whether you’re deep into plasma physics and engineering, skeptical-but–...
266 comments
[ 3.0 ms ] story [ 293 ms ] threadhttps://www.proximafusion.com/press-news/proxima-fusion-and-...
Problems are still many, though (Paper:)
https://www.sciencedirect.com/science/article/pii/S092037962...
I guess the biggest hurdle will be stable operation, without having to replace too many broken parts too often?
Tokamak requires regular shutdown as far as I understood and that is quite a lot of heat stress for all the parts I believe, along with the radiation etc.
(But I lack the background to really debate on the pro vs cons of tokamak vs stellarator, I just have opinions here)
Nuclear aircraft carriers go thirty years without refueling, and fusion could easily do the same.
I can't tell for sure what this means, if it's propulsion or a temporary ship to shore power plant. I suppose this gives them the latitude to target both. Or perhaps the idea is to get exposure to maritime shipping investors.
A mobile power plant might be useful for deep sea mining, or meeting seasonal energy demands for major cities.
Not exactly, but very close to true. As I hear, newest Virginia class of US submarines are planned to be all-electric, but rumors said, they are extremely expensive, so all currently commissioned ones have close to classic propulsion system.
Maritime makes sense from that point of view. However, this point of view ignores that fusion is hard for us right now.
I bet it's a case of someone not understading this field properly and thinking that just throwing money at it will solve everything.
Not really, the energy technologies we've needed have been around for about half a century, with quite reasonable economics (albeit less and less so, as the time pressure increases), especially compared to the alternative. The problems that need to be solved are political and economic, not technical.
You're saying the technical problems involved in fusion power have already been solved. They haven't.
Nope, try reading it again.
Unfortunately, so far exactly 0 working fusion reactors have been build, so currently, I would not demolish electric lines just yet.
Currently they're trying to gas light us that hydrogen-hybrid locomotives are the futures (why not use diesel-hybrid locomotives that already exist is a mistery)
The issue with the US is the distances trains have to go. Mostly short distance trains will be(are?) fully electric but long distance and frieght is diesel series hybrid engines.
And hydrogen comment was about this: https://www.energy.gov/eere/fuelcells/articles/usdot-announc... . My tinfoil hat theory is that every carbon-fuel company pushing for hydrogen knowing well it's energy density isn't enough, so the experiment will fail, and they're going back to diesel saying that green tech isn't ready.
I doubt hydrogen will ever go, because people like me will say "Hindenburg, but already on the ground where it can do the most damage, plus moving at 50-80MPH"
I guess they could keep freight trains real short so they're single engine, but that's still 4 and a half million watts stall. Each one can move itself plus at least 10,000 tonnes (i don't know the conversions, nor care, it's a lot of mass) The world record is 82,000 tonnes, 4.5 miles long, 682 cars, 8 locomotives - in Australia, with american locomotives, hauling iron ore.
I have to ask, are you from or in the US?
GE does make a battery powered locomotive, designed to be used for regenerative braking, it can run at full power by itself for a half hour or so https://en.wikipedia.org/wiki/Wabtec_FLXDrive 3,200,000 watts.
Radiation shielding scales poorly in the downwards direction.
The carrier is nuclear powered and can travel at its top speed indefinitely. But it doesn't except briefly in emergencies, because the rest of the group is powered by oil and would quickly run out of fuel.
Maintenance isn't just about downtime though right? This is gonna have to be supported by your crews traveling globally with trade secret, exotic parts. Not even on the top ten hardest things about this of course.
It's an exciting bet for sure, so good luck - if it works, you're taking a big bite out of a really nasty carbon source.
https://tritiumworkshop.com/products/megaglow-tritium-marker...
So you guys are basically part of the MAGA military complex? Why do you need YC?
Is funding your moonshot concept a way for YC to signal something politically?
I think there is probably some kind of headline bleedover effect where, because of repetition, we start seeing these things everywhere, like the afterimage of a bright light.
In other words: two Cybertruck engineers starting a defense company in February 2025 is naturally going to raise some eyebrows, especially when they're selling a technology that they're counting on other people to invent soon.
More interestingly/generally, your mention of confirmation bias/Bader-Meinhoff raises a question for me: is Y-Combinator really investing more in government and heavy industry as it appears? And it looks like the answer is yes, though it's still a tiny minority compared to general B2B SaaS companies:
https://jaredheyman.medium.com/on-the-last-decade-of-y-combi...
Government is a bigger slicer than it's been since 2017, and similar for industry. More drastically, the pool of companies has greatly constricted in geographic terms, with almost all of the 2024 batch coming from the US. Most importantly--as many of y'all probably already know, but I didn't--they just backed their first pure-defense startup in August, Ares Cruise Missiles.
https://www.ycombinator.com/companies/ares-industries
As far as company mottos go, "Missiles are cool" is a fucking terrifying one... Every day I fear the LessWrong people were less wrong than I thought they were.
YC's process looks something like this: smart founders? check. Technical? check. Big important problem? check. Technically credible on this problem? check. "Did $Thing at $Company" can help on that last point for obvious reasons, but the idea of some triple-bank-shot collusion/corruption with Big Political Players is Too Much Internet.
This matters! I would hate it if any smart, technical, wants-to-work-on-big-problem founder were to read HN threads and think "I don't have connections, so I guess YC is not for me." Please don't anyone think that! If that's you, then YC is for people just like you. You have as good a chance as anyone, precisely because these externalities don't make the difference.
To shift gears to your other question:
> is Y-Combinator really investing more in government and heavy industry as it appears?
I don't know, but it's possible. However, the reason might not be what you'd expect. Trends among YC startups have to do with what founders want to work on, and that is mostly determined by macro factors beyond YC.
I suppose both my points could be summed up like this: "YC is all about the founders". A nice, simple, true sentence, but with a lot of implications that are easy to miss.
I have a small question. Which CAD / 3D / Physics software is used for such design and simulations ?
When you say "Q > 1 within a few (say 3) years" are you talking about your own reactors, or others? For that matter, are you trying to partner with CFS and license their technology or are you intending on starting "from scratch" (from whatever is publicly available)?
If that timeline is for fusion in general, what do you think your timeline looks like? Assuming adequate access to funding how soon can you build a Q>1 reactor? How soon after that can you actually go to market and sell a reactor?
---
On an unrelated note, I'm curious what you think of the current approaches to commercial fusion being attempted. Are Tokamaks the only game in town in your mind? Or do the various other approaches also being tried out right now have a good shot (MIF/Zpinches/etc)? Any particular approaches you think are particularly likely to succeed.
This being ycombinator and a startup I'm obligated to say that I don't ask this question because I think it impacts your commercial viability much, the greatest risks in fusion definitely aren't the competitors. I ask it just because I'm curious what people willing to start a fusion company think of the competitors.
---
Ships make a ton of sense to me as an early market. An 11 figure market (according to my own napkin calculations awhile back) where power is much more expensive than on land. At the same time it's never struck me that the hardest part of building a fusion company is finding a market.
We won’t be the first to Q>1, I’m super excited for SPARC to achieve that and will be prepared with champagne.
We’re targeting early 2030’s for our reactor, we’re going straight for the full thing no sub scale reactor in between (we do have a plan for milestone-ing it out in a meaningful way)
I’ve worked on a few alternative approaches earlier in my career (FRC at Princeton, dense plasma focus at LPP Fusion) … I think all fusion approaches are worth looking at, but I’m placing my chips on the tokamak. If I were to pick a runner up, the stellerator.
Second problem, for space weight is so expensive, that it made fission viable, and solve proliferation problem (very little risk, somebody will steal materials from Moon or beyond).
Although, a good radiator on a spacecraft still seems trivially easy compared to having a working fusion reactor in the first place.
Planned maintenance, sure, but unplanned maintenance means the same lost revenue, plus you're stuck floating in the middle of the pacific ocean, possibly in need of parts or debugging expertise that only exists half a planet away or, for that matter, food. It's certainly a good idea to find a niche to make market entry easier, but I would guess that reliability requirements are actually higher for ships than for microgrids. Find some isolated town or island running off flaky diesel generators on shipped-in fuel and negotiate a reasonable SLA.
This ignores, of course, the bigger problem: making fusion work at all at Q > 1. If it were me, I'd work on solving that before worrying too much about optimizing market entry. So far every single fusion effort has failed to clear that hurdle, and any effort on the other parts is wasted if you can't actually make power.
https://en.wikipedia.org/wiki/Steamboat#Sea-_and_Ocean-going
But if you like sails: my pet hypothetical technology is wind-driven hydrogen tankers (or tankers for some other e-fuel derived from hydrogen) that sail out empty, then cruise around wherever there is plenty of wind. They'd have a turbine/generator setup driven by the water passing by and use the energy harvested there for filling the tank. Cruise around as long as it takes to nearly fill the tank then return to port (and fill the rest on the way back). There's a lot of oceans where systems like that could cruise around on. (same concept could also be used for desalonation, there it would not only be about energy but also about avoiding local brine concentrations)
The largest Q-Max-class gas tanker is 345 meters long [1]. Let's say you manage to fit 3 giant Siemens wind turbines on it, with 100m long blades [2]. It's a bit cramped but let's say you have extenders on the side to make room for all 3 of them. And also let's say you found a way to prevent the ship from tipping over when the wind is strong. By deploying floaters on the side or whatever. Not unsurmountable.
Each of those wind turbine has a rated power of 14.7 MW [2]. Let's say that you found a place where the wind blows super strong (but not too strong) and steady all the time. It's possible, since you are a mobile ship, after all. Let's say that you have a way for the ship to keep in the same place despite the strong and steady wind pushing you constantly. Using engines is going to lower your efficiency, so let's say we found another way.
So, now your ship is generating 45MW constantly. According to ChatGPT, this is 32 kg of hydrogen per second, taking hydrolysis losses into account.
Tanker capacity is 18 620 000 kg of liquid hydrogen. It will take 581 000 seconds to fill up. 9697 minutes, 161 hours, or 6.7 days. Much shorter than I thought... Did I miss something?
[1] https://en.wikipedia.org/wiki/Q-Max [2] https://www.offshorewind.biz/2024/04/22/first-siemens-gamesa...
But I think your maths is wrong somewhere. Hydrogen supplies 33MWh/tonne, and you've stated the ship capacity as 18620 tonnes. 18620/(33*24) gives a generation time of 23 days, even before we allow for hydrolysis overheads.
Marine hydrogen isn't a terrible idea though. Tank weight and bulk is prohibitive for aviation, but less so for shipping.
The ship's capacity in MWh is 18620t x 33MWh/t = 614460MWh.
At 45MW generating capacity, an electric hydrolyser at 80% efficiency delivers hydrogen at a rate of 36MW. That will unfortunately take about 70 years to fill the ship to its maximum capacity.
On a more positive note, 36MW is still a heck of a lot of power, plenty enough to run a mid-sized cruise liner or warship. So a marine generating station with three of these turbines could, for example, refuel a liner once a month, and then that liner have enough fuel to cruise for a month, and so on.
This would require a fuel tank with a more reasonable 750 tonne capacity. That's still several times more than the Shuttle, but not beyond the realms of feasibility - and a stronger, heavier tank allows higher pressures / smaller volumes.
Must start some economics of it. Also marine environment is very unforgiving...
It wouldn't be floating exactly though. Moored and piled into the sea bed, sitting above the waves like rigs and wind turbines.
The mobile hydrolyser (not a boat to solve shipping in a quasi perpetuum mobile away, but an energy harvester that focuses on just that) would solve mooring: the "lateral lift" of the boat would take care of that, just how your plain old America's Cup boat isn't just slowly dragged downwind. It would solve linkup: a serious cost component in not all too conveniently located off shore wind installations is the grid connection. And it would solve intermittency: hydrogen is inconvenient compared to hydrocarbons, but it's super convenient compared to getting even more electricity at a time demand on your grid is already satisfied to saturation.
(GP's math is likely wrong, but the assumption that you could somehow cram multiple turbines from the bigger end of market offerings on a boat and call it a day seems so far off to me that I never really looked at the numbers)
I sure would not expect any returns in days, more like months or years. But if we (humanity) could just solve the purely man-made problem of piracy (or would it technically be salvage?), I believe that a robotic fleet of cruising hydrogenerators could be a huge contribution to our energy needs.
(Give climate change another decade or so and the Arctic Ocean maybe becomes an option, although by then we'll have bigger fish to fry, or perhaps poach).
But yes, fuel-based powerplant, likely as part of a hybrid drive running electric motors powering propellers themselves.
Marine propulsion is already pretty optimised for efficiency (turn on engine, set to cruise power, maintain for 14 days, little acceleration, starts, stops, and/or hills), so a hybrid setup would work pretty effectively.
The far greater challenges are Q>1 and reliable fusion within a ship's structure.
I think it fits squarely in the "requires extraordinary evidence" bucket - what makes you so bold ?
Also, what's you intermediate plans between :
2025 -> Post on HN
2028 -> Q>1 achieved (by you ? by someone else ?)
???? -> ????
20xx -> a ship goes to sea powered by a fusion reactor
???? -> ????
2060 -> fusion is so easy, let's use it for baseload
Sorry if I sound stark, but I'm already burnt out and fed up with the "breakthroughs" on batteries that never materialize - I have a very low tolerance threshold for startups promising fusion for next week ;)
If you're on to something, more power to you - we need that yesterday.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
----------------------------------------------------------------
Dear battery technology claimant,
Thank you for your submission of proposed new revolutionary battery technology. Your new technology claims to be superior to existing lithium-ion technology and is just around the corner from taking over the world. Unfortunately your technology will likely fail, because:
[ ] it is impractical to manufacture at scale.
[ ] it will be too expensive for users.
[ ] it suffers from too few recharge cycles.
[ ] it is incapable of delivering current at sufficient levels.
[ ] it lacks thermal stability at low or high temperatures.
[ ] it lacks the energy density to make it sufficiently portable.
[ ] it has too short of a lifetime.
[ ] its charge rate is too slow.
[ ] its materials are too toxic.
[ ] it is too likely to catch fire or explode.
[ ] it is too minimal of a step forward for anybody to care.
[ ] this was already done 20 years ago and didn't work then.
[ ] by this time it ships li-ion advances will match it.
[ ] your claims are lies
As a corroboration, current hotness in cell phones is existing lithium ion tech with new silicon-carbon anodes.
----------------------------------
Dear Nuclear Fusion Power Claimant
Thank you for your submission of proposed new revolutionary nuclear fusion power technology. Your new technology claims to solve humanity's energy problems, produce unlimited clean energy, and is just months away from commercialization. Unfortunately, your technology will likely fail, because:
[ ] it requires materials that cannot be produced at any scale.
[ ] its energy gain (Q factor) is still substantially less than 1.
[ ] its plasma instabilities modes are completely unknown.
[ ] its plasma modeling behavior relies exclusively on numerical simulations.
[ ] it cannot sustain required plasma confinement criteria
[ ] it cannot handle the neutron flux without rapid degradation of components.
[ ] it requires magnetic fields stronger than currently achievable
[ ] it consumes more energy in cooling systems than it produces.
[ ] your claimed breakthrough violates fundamental physics.
[ ] the same approach was tried in the 1960s, 1970s, 1980s, 1990s and abandoned each time for good reason
[ ] by the time it ships, renewable energy plus storage will be far cheaper.
[ ] your timeline has been "5 years away" for the past 50 years.
[ ] your claims are lies.
Sincerely, The Energy Research Community
2025 -> Post on HN
2028 -> Q>1 achieved (by you ? by someone else ?)
CFS plans Q>1 for 2027 with a tokamak design. If they succeed then there will be plenty of VC for similar designs. I'd place my bets that CFS succeeds with Q>1. And I think the real problem will be the energy flux and neutron handling and thus much more a material sciences problem than a plasma physics problem. Thus the idea to look for a niche that has lower power needs is a very clever one. My bet would be rather on Maritime Fusion than Helion. But nevertheless, CFS will be likely first at Q>1 however there is always space for another competitor.
All of this is driven by HTS. Fusion reactors (generically) scale to the inverse^4 of magnetic field strength. HTS doubled the achievable magnetic field strength of electromagnets, which means that ITER-like performance can be achieved in university-scale reactors at comercially-viable, lower costs.
Dr. Dennis Whyte (MIT Nuclear Eng Prof) gave a great seminar at Berkeley that covered some technical nuances. It's mandatory watching if you want to geek out and understand the fusion hype: https://www.youtube.com/watch?v=rY6U4wB-oYM
Are they going to upgrade it or it’s already obsolete before it was even finished?
Source : I generally don't believe VC-backup startups anymore, but that says more about me than about them (thankfully, sometimes they do stuff, like, 140 chars and useful tools for Russian trolls.)
But let's see how it goes !
I can completely imagine that it goes through stages like
Half a century of painful research at universities -> decades to build multiple prototypes -> years to build a POC -> decades to industrialize the POC -> years to connect the POC to the grid -> ???? -> light bulb moment -> ???? -> profit
I'm not sure we much further than the beginning of step 2. Indeed, (sorry to say that), I'd trust an MIT startup to do that more than a YC startup. But real life will serve as evidence.
But LFP and Sodium Ion are making undeniable progress in cheap usable EV batteries that don't require nickel or cobalt.
I know this is going to sound harsh, but I will personnally consider EV batteries "cheap" only when they pass the very scientific "my mom can use the EV exactly like shes uses her car, and she does not notice" test:
* on one charge, drive 10-50km daily for 350 days in a year
* on one charge, drive 1000km in one go for the holidays
* it must be less than 10kE at a local car dealership
* it can be plugged in your garage
* if you forgot to charge it, you can stop every 10km to recharge in 10m
I know: "it's harsh" ! "We should change our car usage" ! "My mom should live differently to accomodate the technology", etc... Go tell that to my mom.
Still, that's the goal post I've been having for 20y, and 10y ago people where writing headlines about "breakthroughs" that would make this possible any time now. And it's not. So I'm burnt out, kinda.
I'm not sure how old your mom is, but 1000km drives aren't something most people want to be doing past 75 ish. And newer cohorts adapt their lifestyles and expectations to available technology, whether that's charging time, on-demand rental for the occasionally needed longer distance EVs, and so on.
Although I can't help thinking that hybrids are a better fit for this common usage pattern - given a choice between hauling the dead weight of a rarely-used gasoline engine, or the dead weight of the 80% of battery capacity you hardly ever use, the gas engine is cheaper and less demanding of rare metals.
It's a shame that weight considerations mean "rent 80% extra range in a removable module for occasional use" isn't a practical option. You could almost have a 10kWh/50km light EV with a gasoline generator or extra battery in a hitch trailer, but the trade-offs don't quite work.
Alternatively, they will keep living their life exactly the way they have so far, and refuse to use the "new" technology until it has caught up with the "old" one.
"Regulations" could in theory help, if it served as enough of an incentive for car makers to pay some serious R&D and get the "new" tech on par with the old one.
Problem is, in practice, though, the option to "not care about the regulation and elect right wing people until it's dropped" is easier than paying chemist and physicists.
And selling my mom a 2t SUV is much better for the shareholders.
So, never mind. At some point, one "game changing revolutionary breakthrough" will actually break through something and change some game.
Or, my mum will die of old age, bitching about those stupid pseudo "cars" that their grand kids can't use to visit her because the battery is too small.
(No, wait, HN tells my mum will actually not die, thanks to some "game changing AI-powered breakthrough in biotech")
Not sure why you would expect a fusion company to have anything to do with this. The technology is completely different.
(Note: not affiliated with launching company, just a random commenter)
https://x.com/MKVRiscy/status/1893590632901595491
The good news is that neutrons are easily absorbed by steel and water.
The bad news is that the resulting irradiated steel isn't anywhere near as strong as it was pre-irradiation.
https://www.science.org/content/article/fusion-power-may-run...
Did YC ask mostly about feasibility, or were they more interested in the customer's opinion?