Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Skunk Works Reveals Compact Fusion Reactor Details

aviationweek.com

Post reply on HN
Print view View original

341–350 of 375 posts

Re: Skunk Works Reveals Compact Fusion Reactor Details

#341

Post by ntoronto »

tjradcliffe wrote: post #320
logfromblammo wrote: post #241

Earlier quoted context omitted.

Physicists recognize the constraints on development imposed by the laws of physics. Software professionals recognize the constraints on development imposed by entropy, information theory, and other esoteric branches of mathematics. Rather than scoff at this article because Lockheed is making extraordinary claims without the backing of extraordinary evidence, I prefer to scoff because nuclear physics modeling and simu…

Simulations are way easier to write than almost anything else (I'm a nuclear physicist who did this professionally for many years.) The reason why: physics. Or more specifically: conservation laws. Conservation laws are the book-keeping of physics, and they act as global checks on correctness. If your code conserves energy, mass, particle number, charge, momentum, etc it is probably correct in the ways that matter, b…

...and which exercised the weird feature of some x86 processors that had something like 80-bit internal registers but did 64 bit double precision computations, and which allowed the garbage in the extra bits to contaminate the LSB, resulting in very slightly different behaviour in debug and release mode, which gave me fits for weeks...

FWIW, this is a compiler error. Most x86 processors have 80-bit extended floats, which allow 64-bit "pow" to be computed in software with low error (among other benefits). Some compilers "helpfully" compile for 80-bit mode with some combinations of compile options. They should never do this unless it's specifically asked for using a single command-line option, partly for the reason you discovered: it totally messes with replicability. It also messes with implementations of "log1p", Kahan summation, and other high-precision algorithms like double-double arithmetic.

If you ever come across this again, submit a bug report.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#343

Post by hga »

Someone wrote: post #332
hga wrote: post #23

Earlier quoted context omitted.

Compact is good, because until you can fuse helium-3, the reactor is producing neutrons, which will transmute the reactor and its shielding into other, often radioactive elements. So the entire thing will have to be regularly replaced and disposed of (which we can do safely, but too many people believe otherwise).

I do not see how compact matters much here. As far as I understand, the basic reaction used is the same, so that compact design will produce the same number of neutrons per Watt produced as a large one. Wouldn't the effect of its smaller mass be that the container will have to be replaced more often, more or less in the ratio of the masses of the machines?

We really don't know the practical consequences of this systematic transmutation (for all we know, neutron liberating fusion reactors will never be economic). It's entirely possible the device will have to be replaced more often, I'm just pointing out the less mass in it, the less that has to be disposed of when that happens.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#344

Post by drjesusphd »

lovemenot wrote: post #337
drjesusphd wrote: post #321

Earlier quoted context omitted.

It's not mentioned because it's tired, played out, and has no intellectual merit.

Aren't you being rather too hard on fusion there? I heard it could be just a decade or two away.

I was referring to the joke.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#345

Post by crazypyro »

Confusion wrote: post #330
Turing_Machine wrote: post #315

Earlier quoted context omitted.

"Hacker News is full of people from the software field; folks whose professions are often unconstrained by physics" Lockheed Skunk Works, on the other hand, is not. They've been pushing the frontiers of engineering in areas that are very much constrained by physics since 1943, and succeeding. Many times.

What have they successfully designed that was not an airplane?

You realize an airplane is just an insanely complex collection of highly specialized and engineered systems? Just because they specialize in airplanes doesn't mean that the problems they have faced, solved and are still facing today are not relevant to thousands of other engineering problems.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#346

Post by logfromblammo »

tjradcliffe wrote: post #320
logfromblammo wrote: post #241

Earlier quoted context omitted.

Physicists recognize the constraints on development imposed by the laws of physics. Software professionals recognize the constraints on development imposed by entropy, information theory, and other esoteric branches of mathematics. Rather than scoff at this article because Lockheed is making extraordinary claims without the backing of extraordinary evidence, I prefer to scoff because nuclear physics modeling and simu…

Simulations are way easier to write than almost anything else (I'm a nuclear physicist who did this professionally for many years.) The reason why: physics. Or more specifically: conservation laws. Conservation laws are the book-keeping of physics, and they act as global checks on correctness. If your code conserves energy, mass, particle number, charge, momentum, etc it is probably correct in the ways that matter, b…

You are simply skeptical for different reasons than I have.

As a software professional, I have been explicitly instructed to alter a program solely for the purpose of making the output more palatable for customers and investors, at the expense of real-world accuracy. And I did it, because as much as I dislike dishonesty, I also hate searching for new jobs and thoroughly enjoy sleeping under a roof and not starving.

Did the simulation that you wrote have a check for conservation of professional ethics?

You have a great advantage in that subatomic particles are unable to lie to you. Software developers have a capacity for deception exceeding even that of accountants, and we are sometimes asked to use it in unethical ways. One might think that there are reasonable limits, but we still have electronic voting machines that are mysteriously unauditable, and software trading agents programmed to automatically front-run institutional investors.

While the 97% figure was simply made up to mirror the ancestor post, it is possible that all those programs that are not accidentally wrong are intentionally wrong. In your case, you can rule out ill intent for the software you wrote yourself, but as there are potentially hundreds of millions of dollars in funding at stake for this fusion "discovery", I would not discount it for any simulation that suggests this device will work.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#347

Post by peterfirefly »

geon wrote: post #338
peterfirefly wrote: post #331

Earlier quoted context omitted.

> "Top physicists?" Fusion is an engineering problem." So are refrigerators. http://en.wikipedia.org/wiki/Einstein_refrigerator

> It has been suggested that most of the actual inventing was performed by Szilárd, with Einstein merely acting as a consultant and helping with the patent-related paperwork.[1] Plausible, since he had the experience.

And Szilárd was not a top-class physicist (and engineer)?

Re: Skunk Works Reveals Compact Fusion Reactor Details

#348

Post by Confusion »

crazypyro wrote: post #345
Confusion wrote: post #330

Earlier quoted context omitted.

What have they successfully designed that was not an airplane?

You realize an airplane is just an insanely complex collection of highly specialized and engineered systems? Just because they specialize in airplanes doesn't mean that the problems they have faced, solved and are still facing today are not relevant to thousands of other engineering problems.

Sure. The question remains how many of those systems and solutions are relevant to designing a fusion reactor. I think it's easy to overestimate the relevance from a helicopter point of view.

For instance, the same argument applies to a wafer stepper or 'dreadnought class' surface mining equipment. Yet it seems unlikely they are able to commercially compete on both of those extremely different specialized kinds of machines. And I wonder if a fusion reactor would be any different.

Re: Skunk Works Reveals Compact Fusion Reactor Details

#350

Post by aetherson »

jsmthrowaway wrote: post #194
nabla9 wrote: post #164

I'm not seeing any top physicists working in this program (McGuire is not one). I would be willing to bet 1000 EUR that their math and modeling does not add up, or they have skipped some details. Problem with fusion research like this is that the closer you get self sustainment or energy generation, the harder it gets and problems pile up. This project looks like many other similar projects that have gone bust. They…

It's always a safe bet that the top Hacker News comment will always be "won't work, it's too hard." It's like sitting around for a year trying to sell a project at my current employer, with everybody and their lead telling you "that'll never work," then you implement in a weekend and those same people start nitpicking what you've come up with. (What happened to it'll never work? "Well, your CSS isn't great.") Honestl…

Thank you for your detailed explanation of why people should not make any attempt to apply their judgment, experience, or knowledge of the world to press releases, and instead accept them uncritically.
Post reply on HN