Nice. I named our company after the Matrioshka brain in a roundabout way (http://1026labs.com, 10^26 being the wattage of our sun) but we haven't called anything computronium yet.
OT, but the Dyson Sphere was the first idea that gave me a deep sense of sadness knowing I'll never see it.
How can you feel fulfilled laying the groundwork when things like this lay in store? And I'm just a software developer, I can't imagine how people in more relevant roles feel about it.
The other option, which is what I generally feel, is a sense of purpose that we need to make sure we shepherd the human race in the small way we can to the eventual goal of something like this. That is, it's engendered a bit more thoughtfulness about not just what policies yield the best current outcome, but what if we took a path of advancement slightly less optimized for current ease and happiness and more for future gain?
We are of course very far from being able to correctly plan and carry out anything like that if we really wanted to, given that our current advancement is almost entirely directed by market forces and psychology that we are just beginning to understand, but it's an interesting thought experiment.
Can you ethically deny X people currently for the betterment of n*x people in the future, assuming population growth and expansion? If so, what's to keep us from enslaving all the current people in pursuit of ultimate happiness of some theoretical future? If not, why are we saving anything for future generations rather than consuming it all for our benefit right now? The extremes are of course easy to answer, but the variables that go into it make it interesting to think about.
I think that some day the humanity will grow too much and will get tired of overpopulation (I'd hope 30 billion would be the number, but if the humanity is arrogant, they would push it up to 100 billion or more), some people will just die because the planet can't provide enough resources, some rich people will buy other peoples live and kill them (give them a ton of money only if they are willing to die soon) and the world population will agree in not growing anymore but enjoying and preserving the planet as it is, maybe staying in a billion population. After that the world will start to restore and men will have to do nothing because machines are already doing everything, and the only jobs the humanity will have are going to be science and entertainment (incluiding arts and sports)... humanity will no longer grow not because they can't, but because they don't want.
That assumes we are constrained to a single planet, and solar system. I'm not sure why once we have the ability to settle other planets and solar systems, why we would ever really be constrained by overpopulation prior to running up against some other constraint, such as another species.
Well, that's either a (relatively) short-term constraint, or we are no longer advancing. When talking in the context of Dyson spheres, I think that naturally puts the time frame at multiple orders of magnitude longer than any constraint should hold us back, unless we are actually regressing rather than advancing.
Put another way, when talking about tens to hundreds of thousands of years in the future, I'm not sure even a thousand years from now to get even the most rudimentary of interstellar ships registers as more than a blip.
Growth of a space civilization may not be limited by a constant, but by other asymptotes. For example, if FTL travel is really impossible, the accessible space grows with t^3 for three spatial dimensions, which is way less than the exponential growth that or economy is built on.
There obviously will be some constraints on a planet scale, I was addressing more the idea of planet level constraints being a constraint on humanity itself, which implies that humanity is all still on that planet. That is, I think "constraints on the people living on this planet" and "constraints on humanity" won't necessarily always be the same thing, and a in a discussion such as this, a qualification is appropriate.
You are right that the speed of light is very likely a constraint on human population growth based on exactly what you outlined, and while I wasn't thinking of it specifically, falls under what I meant by "some other constraint". :)
Interesting responses. If you haven't already, the Foundation trilogy is a must read. Essentially a bunch of really smart folks create a new field of science focused on predicting the geopolitical and technological advances of the future universe to ensure the human race is successful.
You can see the manned space program. If you're under 43, there's a decent change you'll never see a human leave Earth because you were born too late. Countless humans died before either of these. You'll always be born too early for some things, and too late for others.
What's exciting to me is what lays in store tomorrow, next year, next decade. Next century is someone else's territory, but mine is so vast I have no reason to be jealous of them.
If all the dimming star headlines are piquing your interest, you must read Pandora's Star and Judas Unchained by Alexander Hamilton. It's about the same event being observed but a couple hundred years from now in a post FTL society, so naturally they have to wander out and see what's up. The pair is one of the best recent science fiction stories to come out in the past 15 years.
Yeah, Hamilton really gets into world building. Often to the detriment of actual storytelling -- which on its day, is actually quite good.
I'd still recommend these books though.
Edit to add: Hamilton also wrote The Void Trilogy, set in the same universe as The Commonwealth Saga. They're interesting, but less so than the Commonwealth books.
Recently read it on my phone, which tells me it took 27.4 hours, or about the same as Scott Lynch's first two Gentlemen Bastards novels (also a good read, for different reasons), or a bit longer than the first 4 Dresden Files.
I learned what Dyson sphere is as by product of reading this by Bruce Scheiner, really great read:
" One of the consequences of the second law of thermodynamics is that a certain amount of energy is necessary to represent information. To record a single bit by changing the state of a system requires an amount of energy no less than kT, where T is the absolute temperature of the system and k is the Boltzman constant. (Stick with me; the physics lesson is almost over.)
Given that k = 1.38×10-16 erg/°Kelvin, and that the ambient temperature of the universe is 3.2°Kelvin, an ideal computer running at 3.2°K would consume 4.4×10-16 ergs every time it set or cleared a bit. To run a computer any colder than the cosmic background radiation would require extra energy to run a heat pump.
Now, the annual energy output of our sun is about 1.21×1041 ergs. This is enough to power about 2.7×1056 single bit changes on our ideal computer; enough state changes to put a 187-bit counter through all its values. If we built a Dyson sphere around the sun and captured all its energy for 32 years, without any loss, we could power a computer to count up to 2192. Of course, it wouldn’t have the energy left over to perform any useful calculations with this counter.
But that’s just one star, and a measly one at that. A typical supernova releases something like 1051 ergs. (About a hundred times as much energy would be released in the form of neutrinos, but let them go for now.) If all of this energy could be channeled into a single orgy of computation, a 219-bit counter could be cycled through all of its states.
These numbers have nothing to do with the technology of the devices; they are the maximums that thermodynamics will allow. And they strongly imply that brute-force attacks against 256-bit keys will be unfeasible until computers are built from something other than matter and occupy something other than space."
Your linked version also doesn't mess up the units: 3.2°Kelvin is incorrect, it's 3.2 K. Kelvin is never used with the degree prefix.
I could understand that in copying the quote one might have not noticed the superscripts were messed up, except it is evident deliberate alterations were made. Very strange.
> Your linked version also doesn't mess up the units: 3.2°Kelvin is incorrect, it's 3.2 K. Kelvin is never used with the degree prefix.
This is pointless pedantry. Yes, the official unit is just "Kelvin" now. No, it's not impressive to recognize this misuse. This is little better than pointing out that someone wrote "too" instead of "two".
Historically, it was also called degrees Kelvin, so it's not as if there's a lack of precedent for this, even if it's no longer the preferred usage.
> I could understand that in copying the quote one might have not noticed the superscripts were messed up, except it is evident deliberate alterations were made. Very strange.
It's not evident at all. I don't know why you would assume that some random StackExchange answer is the definitive version. Schneier's original version uses °K, per Scheier's blog:
The argument is that if even just counting through all values of a perfectly energy-efficient 192-bit counter requires the power of an entire star for many years, it's obvious that you'll never have to worry about someone brute-forcing through all keys of something encrypted with 192 bit strength (or better). And even tapping a supernova only gets you 219 bits.
It's about the feasibility of breaking 256-bit encryption keys with a naive brute force attack. The simplest component of a computer that performs this attack would be a 256-bit counter that iterates through all possible key values, however Schneier is saying that according to our current understanding of physics, even if you built the most efficient computer possible and powered it with the energy output of a supernova, it would only have enough energy to iterate through all the values of a 219-bit counter.
The connection between elementary physics and "information" is unproven pop-sci that desperate physicists have come up with in order to stay cool and relevant, in the absence of any real breakthroughs in physics for close to a century.
Throw it in the same rotten bag as string theory and super-symmetry.
That calculation is for irreversible computing. Build your computer out of reversible gates and you can compute using much lower energy (although I believe it will be slower, and need more bits).
I was obsessed with the idea of dyson spheres when I was younger. The are the alternate path of an advanced civilization besides exploring the stars.
You probably wouldn't want to build one around our Sun, since it will eventually expand, and cook anything inside the sphere. Building the sphere would likely take millennia, and there is no reason to invest in building around a start that will eat it. Instead you would build it around a red or white dwarf. These are stable for billions or trillions of years. Meaning you could build a dyson sphere, and live on it until the end of the universe as we know it (Stelliferous era).
Red dwarfs are much less massive than our star, so you could build a smaller sphere (still relatively huge), and material property requirements would be less. White dwarfs are a bit more massive on average, but will last much longer.
Dyson spheres, bubbles, rings, whatever would not experience a day/night cycle. So would either have to build it out of a material that passively radiates excess energy out the other side, while maintaining habitable temperature. Or you build a second set of smaller rings or disks that orbit closer to the star at a different rate. The inner set casts shadows on the outer set, simulating a day/night cycle.
Successfully building a dyson sphere would likely only be done to either power a FTL gate of some sort or a giant computer. If you built a computer to simulate a universe, and properly managed the inner star's output and lifecycle, you could feasibly escape the end of the universe by simulating a smaller one for yourself.
Freeman Dyson actually had another way to escape the heat death of the universe: by taking advantage of it.
As the universe cools, the noise of the background radiation will diminish. Dyson posits that this would allow us to build an analog computer which uses less and less energy as time goes on-- it could run forever on a finite store of energy.
So even if the universe "ends", we might be able to build a computer which doesn't.
If it is possible to build a universe simulator that runs for an infinite amount o time, it's infinitely more likely we live in it than in a real universe.
If you can build a Dyson sphere, you can probably modify a star; reduce and sequester its mass, for instance, so that you can use its energy more efficiently over a longer term. And a star isn't necessarily the most efficient way to burn matter (think "heating your home with a forest fire" when you could have a nice well-designed furnace instead).
An interesting thing about Dyson spheres and red dwarf stars, is that the surface temperature of some red dwarfs is less than the melting temperature of some refractory materials, like tungsten or diamond. In principle, one could build a Dyson sphere directly on the surface of such a star, and use buoyancy or the gas pressure of the star to keep the Dyson Sphere material 'afloat' on the star. Note that the surface gravity of even a dwarf star is very much higher than on a planet; however this is actually helpful if using a buoyant construction.
It's too bad that most red dwarf stars are so low in carbon that extracting the carbon from the star itself to build the sphere may not be feasible. If it were feasible, it would answer the common Dyson Sphere engineering question of "where do you get the mass to build it?".
Another interesting thing is that if you build a Dyson sphere with the intention of using it as a heat engine between the star's surface temperature and the cold of deep space, the Carnot efficiency is so high that the Dyson Sphere's blackbody emission temperature is significantly lower than the star's normal surface temperature. As long as you have an energy sink (such as performing nuclear or chemical reactions that store energy, and leaving it stored), the outer surface of the Dyson Sphere could be a relatively comfortable temperature.
Yep, that was an awesome episode for what it was, and the novel that served as a sequel was great as well.
Unfortunately the version of a Dyson sphere depicted in both is completely untenable, according to Freeman Dyson himself. He reportedly enjoyed the episode as "a piece of cinema" but denounced the science involved as 100% fiction and not possible. As others have said, perhaps building something like that around a dwarf star would be within the realm of possibility, but you still have issues such as radiation, constant daylight, gravity, and drift (the sphere would eventually drift into the star since it's not actually orbiting and is much more massive than even a large planet).
Still, it's a really cool idea and a fun thought experiment.
Exactly, most science fiction representations fail on the engineering test. That' snot to say that some future change in our understanding of physics might not change what is in fact possible. One reason I love scifi is that the authors/readers/watchers are free to drop the need to be fully grounded and go for something truly cool. I suspect what we view as totally impossible today, might be within the realm of possible 2-3k years from now.
If they capture all their sun's energy, shouldn't they get pretty hot?
That's the thing with Tabby's Star. If there really is a Dyson Sphere under construction blocking all that light, shouldn't we be seeing a bump in infrared with the same amount of energy as the blocked light?
Unless their Dyson Sphere-constructing super tech has found another way to get rid of the heat, of course.
If we're talking about powering some future hypothetical civilization.. Yeah it'd have to be ginormous and have huge energy requirements to warrant that much energy. Maybe if they wanted to channel that energy and blow up a few planets it could be useful..
Building the ring would be interesting engineering challenge. Once you build it with sun in its center of mass, it's in unstable equilibrium, so you only need to compensate for the drift. But until it is built, your components have to orbit the sun. So the finished ring would also keep this momentum and it would be rotating around its center.
There are other more plausible options for these civilizations then building absolutely gigantic structures around suns.
1. Build devices/objects (https://i.ytimg.com/vi/bQ7RaOMHb5I/maxresdefault.jpg) that are able to get close to the stars, that would then absorb the emitted energy from the various coronal ejections, magnetic fluxes, etc.
2. There is more energy in the fabric of empty space-time than anywhere else. An advanced civilization would likely be able to harvest this ZPE (zero point energy or field) using rather small devices.
For those who don't know him, he used to be a professional programmer, and he's a very smart writer, lives and his blog is here. http://www.antipope.org/charlie/
I can imagine a Dyson Swarm, although intuitively I wouldn't be surprised if there's some kind of computational limit created by the fact that the bigger a swarm gets, the more computation is needed to keep it gravitationally stable
But I think a Dyson Sphere - something as solid as a planet, but with a radius the size of a planetary orbit - is physically impossible, even with unimagined magic alien super-materials.
The physical configuration of a Dyson sphere is inherently unstable. It's the astronomical equivalent of balancing a pencil on its tip. Any asymmetry at all - including proper motion and radiational asymmetry from the star - would need active correction. Without it, the entire structure will fall apart catastrophically. (It might take a while, but there's not much point building something that big unless it lasts.)
Active correction implies movement, which is not what you want on an inhabitable surface, or even on a computing machine.
And it probably wouldn't be much use as a computing machine. The bigger the system, the longer it takes information to propagate across it. A system with propagation delays on the order of minutes isn't going to run nearly as fast as its energy budget might suggest.
That's still true even if you imagine some kind of ultimate limit system where the computation happens directly on a spherical spacetime shell instead of on tangible hardware.
I suppose it's possible that if it turns out that KIC 8462852 is showing evidence of alien action, the aliens are somehow dimming the star directly instead of by building screens around it.
No. Consider a small perturbation (i.e. Poking) the ball on your table. At least for a small enough poke and/or a big enough table, friction slows down the ball again, and it stops.
Now consider a small poke on a Dyson sphere that was (before the poke) perfectly balanced. The part that got poked is slightly closer to the sun, and the point opposite is slightly further away. So the closer part has some gravitational force acting on it, and the opposite part has a lower gravitational force acting on it: O(1/(r-e)^2) vs O(1/(r+e)^2). The imbalance causes a net gravitational force in the direction of the poke, which eventually crashes into the sun.
In fictional accounts, the Dyson-sphere concept is often interpreted as an artificial hollow sphere of matter around a star. This perception is based on a literal interpretation of Dyson's original short paper introducing the concept. In response to letters prompted by this paper, Dyson replied, "A solid shell or ring surrounding a star is mechanically impossible. The form of 'biosphere' which I envisaged consists of a loose collection or swarm of objects traveling on independent orbits around the star."
As I see it, "Dyson sphere" designates the broader category of such devices. You're saying a literal Dyson sphere would be in a perfectly metastable configuration, which is true, though I suppose a solid structure resembling a sphere might be engineered to be in a stable configuration and have some tolerance to perturbation (unlikely though it may be).
And in the case of a computer, there's no reason to assume that the processing would take place all over the sphere, which is just an energy harvesting tool. The energy would be directed to wherever it's needed, for example an AMD graphics card. I also like the idea of dimming the star directly, though I can't begin to imagine the processes involved. Or rather, I can begin to imagine them but I can't come up with something that holds together even remotely.
Seems to me that if you have the energy output of an entire star, you have quite enough energy to keep your shell balanced around the center of it.
Even with those magic reaction-less engines they are working on now you could keep the thing in place by just spacing engines on the outside at the correct distance.
I imagine the movement would have to be insanely slow to prevent the entire structure from collapsing though, and I wonder whether a solar flare wouldn't put out so much energy/movement that you'd spent years correcting it.
A variation on this and the first rung of he Dyson swarm would be a space station that follows the earth's orbit rather than orbits it. you could call it a dyson tile or fly or something like that.
One issue of ring worlds is traction (and Dyson spheres are much more complicated than that). Nothing we can imagine and certainly nothing we can build can take it but what if instead of rotating it above orbital speed to create gravity inside the ring we rotate it below orbital speed, create gravity on the outer surface and rely on materials that can withstand the compression?
Sunlight could still be reflected by a fleet of orbiting mirrors that wouldn't orbit the star in sync with the ring and thus would not be subjected to the structural forces of the ring. In case the ring is built around a smaller star with a spectral emission very different from the one preferred by the ring builders, the mirrors could adjust that too.
When the ring rotates at orbital speed, there is no perceived gravity on either side. If the ring rotates faster than that, one would perceive a force pulling them away from the central star. If, however, the ring rotates below orbital speed its mass would be pulled toward the central star (along with anyone standing on its outer surface). The forces pulling it towards the star would compress its structure (the opposite of the traction forces a faster ring with positive gravity on the inner surface would experience).
I think (I haven't checked the numbers) the ring's mass can't be neglected, but, because of the shape, it'd act as an extra mass within the central star. The ring is, in fact, in its own orbit.
Good and bad news. A stationary ring around a star with one solar mass would be very close, about a tenth of the distance Mercury orbits the sun, to make its inhabitants experience one Earth gravity on its outer surface. The good news is the ring would be much smaller than 1AU and it'd be easier to do it around a large white dwarf and still get a lot of energy from it. Not sure how much pressure solar wind would be, but, if the star is active enough, we could use it to partly sustain the ring structure.
88 comments
[ 3.4 ms ] story [ 172 ms ] threadhttps://en.wikipedia.org/wiki/Computronium
Edit: We're on the way... http://www.theregister.co.uk/2016/07/25/semiconductor_indust...
How can you feel fulfilled laying the groundwork when things like this lay in store? And I'm just a software developer, I can't imagine how people in more relevant roles feel about it.
We are of course very far from being able to correctly plan and carry out anything like that if we really wanted to, given that our current advancement is almost entirely directed by market forces and psychology that we are just beginning to understand, but it's an interesting thought experiment.
Can you ethically deny X people currently for the betterment of n*x people in the future, assuming population growth and expansion? If so, what's to keep us from enslaving all the current people in pursuit of ultimate happiness of some theoretical future? If not, why are we saving anything for future generations rather than consuming it all for our benefit right now? The extremes are of course easy to answer, but the variables that go into it make it interesting to think about.
Put another way, when talking about tens to hundreds of thousands of years in the future, I'm not sure even a thousand years from now to get even the most rudimentary of interstellar ships registers as more than a blip.
You are right that the speed of light is very likely a constraint on human population growth based on exactly what you outlined, and while I wasn't thinking of it specifically, falls under what I meant by "some other constraint". :)
What makes you so certain you will never live to see it? Be proactive and move your efforts to life extension and dependent technology.
What's exciting to me is what lays in store tomorrow, next year, next decade. Next century is someone else's territory, but mine is so vast I have no reason to be jealous of them.
Hopefully Elon Musk will come through.
Pandora's Star is an excellent (albeit long) read.
Note however, the author is Peter F. Hamilton.
I'd still recommend these books though.
Edit to add: Hamilton also wrote The Void Trilogy, set in the same universe as The Commonwealth Saga. They're interesting, but less so than the Commonwealth books.
" One of the consequences of the second law of thermodynamics is that a certain amount of energy is necessary to represent information. To record a single bit by changing the state of a system requires an amount of energy no less than kT, where T is the absolute temperature of the system and k is the Boltzman constant. (Stick with me; the physics lesson is almost over.)
Given that k = 1.38×10-16 erg/°Kelvin, and that the ambient temperature of the universe is 3.2°Kelvin, an ideal computer running at 3.2°K would consume 4.4×10-16 ergs every time it set or cleared a bit. To run a computer any colder than the cosmic background radiation would require extra energy to run a heat pump.
Now, the annual energy output of our sun is about 1.21×1041 ergs. This is enough to power about 2.7×1056 single bit changes on our ideal computer; enough state changes to put a 187-bit counter through all its values. If we built a Dyson sphere around the sun and captured all its energy for 32 years, without any loss, we could power a computer to count up to 2192. Of course, it wouldn’t have the energy left over to perform any useful calculations with this counter.
But that’s just one star, and a measly one at that. A typical supernova releases something like 1051 ergs. (About a hundred times as much energy would be released in the form of neutrinos, but let them go for now.) If all of this energy could be channeled into a single orgy of computation, a 219-bit counter could be cycled through all of its states.
These numbers have nothing to do with the technology of the devices; they are the maximums that thermodynamics will allow. And they strongly imply that brute-force attacks against 256-bit keys will be unfeasible until computers are built from something other than matter and occupy something other than space."
[1] https://en.wikipedia.org/wiki/Gray_code
That's ignoring all the logic required to figure out exactly which bits to flip...
I could understand that in copying the quote one might have not noticed the superscripts were messed up, except it is evident deliberate alterations were made. Very strange.
This is pointless pedantry. Yes, the official unit is just "Kelvin" now. No, it's not impressive to recognize this misuse. This is little better than pointing out that someone wrote "too" instead of "two".
Historically, it was also called degrees Kelvin, so it's not as if there's a lack of precedent for this, even if it's no longer the preferred usage.
> I could understand that in copying the quote one might have not noticed the superscripts were messed up, except it is evident deliberate alterations were made. Very strange.
It's not evident at all. I don't know why you would assume that some random StackExchange answer is the definitive version. Schneier's original version uses °K, per Scheier's blog:
https://www.schneier.com/blog/archives/2009/09/the_doghouse_...
As does the original answer on StackExchange, actually (look at the revision 4 diff):
http://security.stackexchange.com/posts/25392/revisions
Actually, it is.
It's a geek pedantry, the best kind there is :)
I was wondering what's so special about having a computer count to 2192 and why that would require so much energy.
Throw it in the same rotten bag as string theory and super-symmetry.
Solar operated mirrors whose focus and intensity can be regulated remotely.
You probably wouldn't want to build one around our Sun, since it will eventually expand, and cook anything inside the sphere. Building the sphere would likely take millennia, and there is no reason to invest in building around a start that will eat it. Instead you would build it around a red or white dwarf. These are stable for billions or trillions of years. Meaning you could build a dyson sphere, and live on it until the end of the universe as we know it (Stelliferous era).
Red dwarfs are much less massive than our star, so you could build a smaller sphere (still relatively huge), and material property requirements would be less. White dwarfs are a bit more massive on average, but will last much longer.
Dyson spheres, bubbles, rings, whatever would not experience a day/night cycle. So would either have to build it out of a material that passively radiates excess energy out the other side, while maintaining habitable temperature. Or you build a second set of smaller rings or disks that orbit closer to the star at a different rate. The inner set casts shadows on the outer set, simulating a day/night cycle.
Successfully building a dyson sphere would likely only be done to either power a FTL gate of some sort or a giant computer. If you built a computer to simulate a universe, and properly managed the inner star's output and lifecycle, you could feasibly escape the end of the universe by simulating a smaller one for yourself.
As the universe cools, the noise of the background radiation will diminish. Dyson posits that this would allow us to build an analog computer which uses less and less energy as time goes on-- it could run forever on a finite store of energy.
So even if the universe "ends", we might be able to build a computer which doesn't.
[1] https://www.amazon.com/Last-Three-Minutes-Conjectures-Ultima...
It's too bad that most red dwarf stars are so low in carbon that extracting the carbon from the star itself to build the sphere may not be feasible. If it were feasible, it would answer the common Dyson Sphere engineering question of "where do you get the mass to build it?".
Another interesting thing is that if you build a Dyson sphere with the intention of using it as a heat engine between the star's surface temperature and the cold of deep space, the Carnot efficiency is so high that the Dyson Sphere's blackbody emission temperature is significantly lower than the star's normal surface temperature. As long as you have an energy sink (such as performing nuclear or chemical reactions that store energy, and leaving it stored), the outer surface of the Dyson Sphere could be a relatively comfortable temperature.
Unfortunately the version of a Dyson sphere depicted in both is completely untenable, according to Freeman Dyson himself. He reportedly enjoyed the episode as "a piece of cinema" but denounced the science involved as 100% fiction and not possible. As others have said, perhaps building something like that around a dwarf star would be within the realm of possibility, but you still have issues such as radiation, constant daylight, gravity, and drift (the sphere would eventually drift into the star since it's not actually orbiting and is much more massive than even a large planet).
Still, it's a really cool idea and a fun thought experiment.
If they capture all their sun's energy, shouldn't they get pretty hot?
That's the thing with Tabby's Star. If there really is a Dyson Sphere under construction blocking all that light, shouldn't we be seeing a bump in infrared with the same amount of energy as the blocked light?
Unless their Dyson Sphere-constructing super tech has found another way to get rid of the heat, of course.
If we're talking about providing for the energy needs of the Earth, that'd be way way overkill too. See how much solar we'd need on Earth to power our needs: http://www.techinsider.io/map-shows-solar-panels-to-power-th...
If we're talking about powering some future hypothetical civilization.. Yeah it'd have to be ginormous and have huge energy requirements to warrant that much energy. Maybe if they wanted to channel that energy and blow up a few planets it could be useful..
Demand follows supply.
1. Build devices/objects (https://i.ytimg.com/vi/bQ7RaOMHb5I/maxresdefault.jpg) that are able to get close to the stars, that would then absorb the emitted energy from the various coronal ejections, magnetic fluxes, etc.
2. There is more energy in the fabric of empty space-time than anywhere else. An advanced civilization would likely be able to harvest this ZPE (zero point energy or field) using rather small devices.
For those who don't know him, he used to be a professional programmer, and he's a very smart writer, lives and his blog is here. http://www.antipope.org/charlie/
But I think a Dyson Sphere - something as solid as a planet, but with a radius the size of a planetary orbit - is physically impossible, even with unimagined magic alien super-materials.
The physical configuration of a Dyson sphere is inherently unstable. It's the astronomical equivalent of balancing a pencil on its tip. Any asymmetry at all - including proper motion and radiational asymmetry from the star - would need active correction. Without it, the entire structure will fall apart catastrophically. (It might take a while, but there's not much point building something that big unless it lasts.)
Active correction implies movement, which is not what you want on an inhabitable surface, or even on a computing machine.
And it probably wouldn't be much use as a computing machine. The bigger the system, the longer it takes information to propagate across it. A system with propagation delays on the order of minutes isn't going to run nearly as fast as its energy budget might suggest.
That's still true even if you imagine some kind of ultimate limit system where the computation happens directly on a spherical spacetime shell instead of on tangible hardware.
I suppose it's possible that if it turns out that KIC 8462852 is showing evidence of alien action, the aliens are somehow dimming the star directly instead of by building screens around it.
No, it's neutral--the equivalent of keeping a ball on a table.
Now consider a small poke on a Dyson sphere that was (before the poke) perfectly balanced. The part that got poked is slightly closer to the sun, and the point opposite is slightly further away. So the closer part has some gravitational force acting on it, and the opposite part has a lower gravitational force acting on it: O(1/(r-e)^2) vs O(1/(r+e)^2). The imbalance causes a net gravitational force in the direction of the poke, which eventually crashes into the sun.
In fictional accounts, the Dyson-sphere concept is often interpreted as an artificial hollow sphere of matter around a star. This perception is based on a literal interpretation of Dyson's original short paper introducing the concept. In response to letters prompted by this paper, Dyson replied, "A solid shell or ring surrounding a star is mechanically impossible. The form of 'biosphere' which I envisaged consists of a loose collection or swarm of objects traveling on independent orbits around the star."
As I see it, "Dyson sphere" designates the broader category of such devices. You're saying a literal Dyson sphere would be in a perfectly metastable configuration, which is true, though I suppose a solid structure resembling a sphere might be engineered to be in a stable configuration and have some tolerance to perturbation (unlikely though it may be).
https://en.wikipedia.org/wiki/Metastability
And in the case of a computer, there's no reason to assume that the processing would take place all over the sphere, which is just an energy harvesting tool. The energy would be directed to wherever it's needed, for example an AMD graphics card. I also like the idea of dimming the star directly, though I can't begin to imagine the processes involved. Or rather, I can begin to imagine them but I can't come up with something that holds together even remotely.
Even with those magic reaction-less engines they are working on now you could keep the thing in place by just spacing engines on the outside at the correct distance.
I imagine the movement would have to be insanely slow to prevent the entire structure from collapsing though, and I wonder whether a solar flare wouldn't put out so much energy/movement that you'd spent years correcting it.
Sunlight could still be reflected by a fleet of orbiting mirrors that wouldn't orbit the star in sync with the ring and thus would not be subjected to the structural forces of the ring. In case the ring is built around a smaller star with a spectral emission very different from the one preferred by the ring builders, the mirrors could adjust that too.
That's a tricky one. Do you mean make the ring thick enough that it has enough mass to have its own gravity?
When the ring rotates at orbital speed, there is no perceived gravity on either side. If the ring rotates faster than that, one would perceive a force pulling them away from the central star. If, however, the ring rotates below orbital speed its mass would be pulled toward the central star (along with anyone standing on its outer surface). The forces pulling it towards the star would compress its structure (the opposite of the traction forces a faster ring with positive gravity on the inner surface would experience).
I think (I haven't checked the numbers) the ring's mass can't be neglected, but, because of the shape, it'd act as an extra mass within the central star. The ring is, in fact, in its own orbit.