back
4 comments
Gain over incident laser energy, but not over total input energy. Still two orders of magnitude to go, based on 300MJ inputs. Not gonna happen.

Laser inertial confinement is a commercial dead end. It's only continued as an inertial funding exercise, reusing nuclear weapon simulation facilities long after supercomputers took over the modelling.

Note that 10 years ago this facility was maxing out at 14.4 kJ, now they are at 3.5 MJ, well over a two order of magnitude improvement. As you can see in [0] (which doesn't even include the latest results) their progress has thus far been exponential. While this facility is not designed to achieve engineering breakeven, it is designed to get past the 100 MJ level.

[0] https://en.wikipedia.org/wiki/National_Ignition_Facility#/me...

Two orders of magnitude to go relative to old inefficient lasers used at the facility. That said, I agree that this kind of laser inertial confinement will likely not produce a power source. It's still valuable though as it verifies the achievability of fusion ignition and the conditions needed.

Furthermore, supercomputers do not replace these facilities, only complement them, as modeling a system solely in computers is a good way to find you have drifted away from reality.

> It's only continued as an inertial funding exercise, reusing nuclear weapon simulation facilities long after supercomputers took over the modelling.

Sure, but that's also why they don't care about the inefficiency of the lasers themselves, which we already know how to improve upon significantly.

Which means…

> Still two orders of magnitude to go, based on 300MJ inputs. Not gonna happen.

Might happen.

Not guaranteed — even good lasers aren't great, though apparently 60.5% wall-plug efficiencies have been demonstrated[0], and even if the laser part was perfect there would be a lot more steps from this to making a power station (or a fusion rocket) — but it's certainly in the realm of being worth considering.

[0] still a loss overall, assuming 150% gain from laser in to fusion out: https://ui.adsabs.harvard.edu/abs/2020JSemi..41c2901R/abstra...