What you're missing is that these pictures aren't of single nanotubes, but rather of strips or strands consisting of numerous nanotubes, and they're not intended as microprocessor components at all. Nanotubes have potential applications all over, not just in microelectronics.
These macroscopic fibers allow macro-scale experimentation on the properties of nano-tubes. While they are too large to serve as, say, interconnects on microchips, they have a host of other uses. While much is hypothesized about the properties of nanotubes, not much is known concretely. Hopefully, these will allow experiments to determine the actual physical properties of the tubes.
The article also mentions using the nanotube fibers as a substite for existing carbon fiber components, which would allow these to become even lighter and stronger, and at first glance would seem to solve the problems some of these materials have with brittleness and low tolerances for linear stresses.
So in essence, what you want is a tiny little harddisk, with a tiny little amount of space on it, drawing huge wattage. I don't see what makes a harddisk so infinitely superior to flash memory, embedded or cartridge style. Flash memory, and other embedded memory solutions, have speeds and storage capacities rivalling those of your tiny drive, but consume many times less power. Solid state memory is also much much more durable than that tiny drive you like so much. In the future, PDAs will most probably be expected to take even more abuse than they do today. A tiny magnetic platter drive won't take much. Heck, Western Digital even recommends against placing the harddisk on a flat table without padding. Magnetic platter harddisks aren't the future. If you're waiting until PDAs come with minidrives, i would have to advise against holding your breath. Unless someone finds a market for a large, heavy, noisy, fragile power hog of a PDA. Sound likely?
The Einsteinian mass-energy relation does indeed apply to hydrocarbon combustion. If one takes the reaction CH4+2O2-->CO2+2H20 to complete combustion in a sealed system, allowing the resulting heat to escape to the surroudings, the mass of the final system will be infinitesimally smaller than the mass of the initial. As I am given to understand it, the energy of the chemical bonds exists as a component of the mass of the object. Or possibly the other way around. I forget that particular lecture. (keep in mind that i am merely an undergrad physicist, and don't really know much at all) Thus, the heat released in combustion comes from the conversion of a infinistesimally tiny fraction of the mass to energy. But then, i could be entirely wrong.
Is this a typo, or is this stuff actually that full of energy?
The Helium-3 atom is not in and of itself drastically energetic. One could not fill a bomb with He-3 alone and expect it to do anything more drastic than make the target talk funny. The excitment expressed in this article is due to the high degree of effeciency of a Fusion reaction, which can be attained with He-3. Traditional energy sources, such as hydrocarbon combustion, only convert a tiny percentage of the available mass to energy. As shown by the famous equation E=mc^2, even a tiny amount of mass converted to energy results in quite a bit of energy. If you doubt that, go light a can of gasoline on fire. It's hot. In a comprable He-3 fusion reaction, a much greater (but stil small) fraction of the total mass is converted into energy. Thus, a small amount of He-3 could theoretically provide the same power yeild of a much greater amount of gasoline, or natural gas.
How long until a distributed nuclear simulation project? I guess that wouldn't happen becuase of "security concerns," though. originally posted in response to a thread in the first, and now deleted, item. The collective community here has been over this scores of times already. To reiterate: Seti@Home, all of the various key breaking projects, and any other distributed processing project all have one thing in common: the data chunk being processed by one computer is largely independent of the other data chunks being processed by the others. Hence, it is feasible to send a chunk out from the central distributor, have it processed by the proccessing computer, and then return it directly to the processing computer. In this application, however, this is just not a possibility. Consider the simulations this computer will be used for, nano-second scale simulations of the initial criticality of a chunk of uranium or plutonium. This is just not something that can be broken up into chunks and handed out for processing. For the simulation to be even vaguely useful, each atom or particle of fissionable substance simulated must interact with all of the other atoms in the simulation. Thus, if someone were to be given a chunk of, say, a million atoms/particle/lattice points, to process, their sub-simulation would need to communicate continuously with all of the other processors sub-simulations. If we look at the system conservatively, and say that each particle will only be affected by the particles closer too it, this still requires the exchange of vast quantities of information at each step. And that's not even taking into account the effects of simulating the neutrons and other sub-atomic particles zipping around. The processors would have to have huge amounts of data transfer ability to make this feasible. Of course, if i misread your intent, and you're actually suggesting that every citizen should be given this huge amounts of banwidth to enable the distributed processing, i'm behind you all the way.
Slashdot Mirror
These macroscopic fibers allow macro-scale experimentation on the properties of nano-tubes. While they are too large to serve as, say, interconnects on microchips, they have a host of other uses. While much is hypothesized about the properties of nanotubes, not much is known concretely. Hopefully, these will allow experiments to determine the actual physical properties of the tubes.
The article also mentions using the nanotube fibers as a substite for existing carbon fiber components, which would allow these to become even lighter and stronger, and at first glance would seem to solve the problems some of these materials have with brittleness and low tolerances for linear stresses.
So in essence, what you want is a tiny little harddisk, with a tiny little amount of space on it, drawing huge wattage. I don't see what makes a harddisk so infinitely superior to flash memory, embedded or cartridge style. Flash memory, and other embedded memory solutions, have speeds and storage capacities rivalling those of your tiny drive, but consume many times less power. Solid state memory is also much much more durable than that tiny drive you like so much. In the future, PDAs will most probably be expected to take even more abuse than they do today. A tiny magnetic platter drive won't take much. Heck, Western Digital even recommends against placing the harddisk on a flat table without padding. Magnetic platter harddisks aren't the future. If you're waiting until PDAs come with minidrives, i would have to advise against holding your breath. Unless someone finds a market for a large, heavy, noisy, fragile power hog of a PDA. Sound likely?
The Einsteinian mass-energy relation does indeed apply to hydrocarbon combustion. If one takes the reaction
CH4+2O2-->CO2+2H20
to complete combustion in a sealed system, allowing the resulting heat to escape to the surroudings, the mass of the final system will be infinitesimally smaller than the mass of the initial. As I am given to understand it, the energy of the chemical bonds exists as a component of the mass of the object. Or possibly the other way around. I forget that particular lecture. (keep in mind that i am merely an undergrad physicist, and don't really know much at all) Thus, the heat released in combustion comes from the conversion of a infinistesimally tiny fraction of the mass to energy.
But then, i could be entirely wrong.
Is this a typo, or is this stuff actually that full of energy?
The Helium-3 atom is not in and of itself drastically energetic. One could not fill a bomb with He-3 alone and expect it to do anything more drastic than make the target talk funny. The excitment expressed in this article is due to the high degree of effeciency of a Fusion reaction, which can be attained with He-3.
Traditional energy sources, such as hydrocarbon combustion, only convert a tiny percentage of the available mass to energy. As shown by the famous equation E=mc^2, even a tiny amount of mass converted to energy results in quite a bit of energy. If you doubt that, go light a can of gasoline on fire. It's hot. In a comprable He-3 fusion reaction, a much greater (but stil small) fraction of the total mass is converted into energy. Thus, a small amount of He-3 could theoretically provide the same power yeild of a much greater amount of gasoline, or natural gas.
How long until a distributed nuclear simulation project? I guess that wouldn't happen becuase of "security concerns," though. originally posted in response to a thread in the first, and now deleted, item. The collective community here has been over this scores of times already. To reiterate: Seti@Home, all of the various key breaking projects, and any other distributed processing project all have one thing in common: the data chunk being processed by one computer is largely independent of the other data chunks being processed by the others. Hence, it is feasible to send a chunk out from the central distributor, have it processed by the proccessing computer, and then return it directly to the processing computer.
In this application, however, this is just not a possibility. Consider the simulations this computer will be used for, nano-second scale simulations of the initial criticality of a chunk of uranium or plutonium. This is just not something that can be broken up into chunks and handed out for processing. For the simulation to be even vaguely useful, each atom or particle of fissionable substance simulated must interact with all of the other atoms in the simulation. Thus, if someone were to be given a chunk of, say, a million atoms/particle/lattice points, to process, their sub-simulation would need to communicate continuously with all of the other processors sub-simulations. If we look at the system conservatively, and say that each particle will only be affected by the particles closer too it, this still requires the exchange of vast quantities of information at each step. And that's not even taking into account the effects of simulating the neutrons and other sub-atomic particles zipping around. The processors would have to have huge amounts of data transfer ability to make this feasible.
Of course, if i misread your intent, and you're actually suggesting that every citizen should be given this huge amounts of banwidth to enable the distributed processing, i'm behind you all the way.