Clever idea. The gadget apparently uses a pulsed solid-state UV laser to ionize a channel in the air between the shooter and the target. The plasma is a much lower resistance path than the un-ionized air, and so the discharge from the gun follows the plasma to the target and then to ground. Interestingly, at high enough intensities laser pulses like that can be self-focusing (pdf) .
Of course, you still need to hike around a whopping big capacitor bank to have this work over any reasonable distance, and the repeat rate of fire would probably be lousy since the capacitors would have to be recharged....
The whole coolness of quantum computation lies in the idea of superposition. The input quantum bits don't just have to be zero or one, but instead can be in a superposition of zero and one. This is powerful in two ways.
First, in principle you can prepare a superposition of all possible inputs to your program. Run the program once. You've now got a superposition of all possible outputs that can be generated from your inputs.
Second, within the program itself, performing an operation some number of times N can lead to superpositions containing ~ exp(N) terms. That is, with a linear number of operations you can generate an exponentially large number of states.
You can read more about quantum computation here or here .
Now if you got the degree through academic malfeasance, that's a different matter -- but I checked the article and all of this guy's sins seem to have been post-graduation.
As part of my own research, I'm very very familiar with the Schon investigation. There are definitely questions about his PhD work. I would be willing to wager a large sum of money that the university wrote it up this way to save face, rather than admit that Schon snowed them and received a PhD for faked data.
First, my bone fides: I'm actively involved in nanoscale research. I've got a PhD and postdoc experience in experimental physics, and run a research group working on this stuff.
There is no way on earth that there will be a general purpose 'fabricator' within twenty years. I'm not saying that I wouldn't like one - it's just not physically realistic!
The founders of this little group have no technical background in this field at all. Through sheer chutzpah, they've declared themselves to be authorities on nanotechnology because they've read popular treatments like those by Drexler. I'm really sick of people with no real scientific credibility becoming pundits merely by being loud.
Unless you actually understand the technology, you can't assess it reasonably! . How hard is this to figure out?!
What kind of search engine turns up no links (none, zero, zip, nada) when searching for "bell labs" ?
I assume their server is aflame or something.
Space as a research environment
on
Bionic Eyes
·
· Score: 1
Those who've read the article know that the material used in these devices is produced by a crystal growth process called molecular beam epitaxy. Growing high purity crystals this way requires "ultra-high vacuums" - that is, background pressures 10^15 times lower than atomospheric pressure.
Dr. Ignatiev at the University of Houston is in charge of a NASA-funded effort to develop "Space Vacuum Epitaxy", where researchers try to take advantage of the vacuum of space as a growth environment. The problem is, near-earth space simply isn't low enough pressure to work well, meaning that the researchers need to have their experiments behind a "wake-shield" - basically a big metal plate that pushes unwanted gas molecules out of the way.
The upshot is that space vacuum epitaxy is hugely expensive. Better and more controlled vacuums may be achieved on earth at a tiny fraction of the cost of a single shuttle launch. While I'm glad NASA funds research efforts like this, many such projects, even while producing cool results, often end up with the conclusion that doing the work in space is neither necessary nor desirable.
I'm a solid state physicist, too, and I've taken classes from Laughlin and Zhang, and met both Pines and Gross.
Never trust an physics article with the words "synergy", "epistemological", or "paradigm" in either the title or the abstract. They're universally crap. If they weren't crap, the author wouldn't need to obfuscate the reader by using flowery information-free phrases.
Umm, the article was quoting a Nobel laureate.
If Laughlin wants to use those words, he has the right.
Second, solid state physics types don't try to study relativity for two reasons: first, we don't have the background to do so credibly and second, general relativity is not observable in the laboratory and thus is not a subject most solid state physicists are interested in.
Zhang started out as a high energy theorist (and a successful one at that) before getting interested in condensed matter. He's more than mathematically capable of doing GR.
Solid state physicists (or condensed matter physicists, same idea) do not specialize in "how emergent pheomena occur" (whatever the hell that means). We study matter in the solid state (superconductors or semiconductor physics for example). Touchy-feely stuff, things you can put on a lab bench.
Now you're just being peevish. Any CM person studying correlated electronic systems, for example, like doing transport measurements in dilute 2d electron gases in GaAs, is certainly interested in how emergent phenomena occur. Wouldn't you say that lots of CM physicists are interested in the emergence of the collective superconducting state from a (strange) metallic normal state in high-Tc superconductors? Anderson was right - more is different.
Just because I can't sit down with a pen and paper (or a Cray supercomputer) and solve the resulting system of equations doesn't mean they don't apply. These so-called "new laws" are just approximations to the real (albeit unwieldy) quantum mechanics.
The language in the article is poorly worded here. However, your point, while strictly true, is not very useful. Come on - we could describe the motion of all the air molecules in this room using Newton's laws, but isn't the collective language of statistical mechanics (pressure, temperature, chem. potential) more useful? Your reductionist point of view is the antithesis of statistical physics that you claim to understand.
The difference is, quasi-particles are not real. They're a useful mathematical trick, but without an underlying object (usually a crystal of some kind), they wouldn't exist. Electronts are real.
Give me a definition of "particle" that makes this self-evident. In quantum field theory, electrons are just as real as phonons. They're quantized excitations of a quantum field.
For someone with a PhD in soft CM, your viewpoints seem pretty unusual. Zhang and Laughlin are two of the smartest people around, and their ideas have a lot of appeal. Isn't it amazing that, out of 10^23 strongly interacting particles in 1 cc of copper, we still end up with excitations (quasiparticles) that act a lot like free electrons? Why are you so deeply offended by the idea that the excitations we are familiar with are actually statistical, collective properties of some other fields? Is that any more ugly than postulating the standard model, complete with its weird, CP-violating lagrangian?
The actual scientific paper is only available through Science magazine's online section right now.
That's subscription only, unfortunately. The print version will appear in Science in a few weeks.
The reason there's no preprint circulating is that Science (and Nature) are notoriously draconian about that sort of thing. Papers are embargoed until the date of publication.
A real paper does exist, though - I've got it right in front of me.
Gain is a common figure of merit of transistor-based amplifier circuits. The gain of a voltage amplifier is defined as the ratio of the size of the output signal to the size of the input signal. An amplifier that could take a 0.5V amplitude sine wave as its input and produce a 5V amplitude sine wave as its output has a gain of 10. You don't get something for nothing, of course - the amplifier has to be connected to an external power source.
A transistor is a three-terminal device. In a typical computer chip, these three terminals are called the source, the drain, and the gate. For a given voltage between the source and drain, the current that flows into the drain is strongly dependent on the voltage applied to the gate. That's what allows transistors to be used as switches: you can make a transistor that won't let current flow from source to drain unless the gate voltage is turned up past some value.
Achieving actual gain in a single-molecule device is important. Without gain greater than one, it's not possible to efficiently chain large numbers of transistors together to manipulate signals. A strong input would get degraded with each stage of transistor manipulation, eventually falling to a level too small to drive subsequent transistors.
There are *many* problems with the idea of using individual molecules to replace Si devices. Achieving a gain > 1 is a necessary but by no means sufficient step for eventual molecule-based computers. As a physicist, I think it's important to recognize real achievements in this field, but not to buy into the hype unquestioningly.
E-beam lithography is not new. This press
release contained frustratingly little actual
information. *All* e-beam machines (except
projection systems like SCALPEL ) raster
an electron beam. That's why e-beam lithography
is historically slow - it's a serial process.
I've done extensive nanofabrication, and these
guys have chosen their words so carefully as
to be misleading. When they talk about making
structures on the "subcellular" scale for
biological research, it sounds impressive
but really isn't. A typical red blood cell
is 5 microns across. The smallest features
produced photolithographically for your Athlon
are 0.13 microns across. Even more annoying
is their claim of molecular and submolecular
scale device size without actually naming a
number. Molecules can be big - DNA can be many
microns long when uncoiled.
A meaningful
figure of merit for resolution is: how small
a feature can you pattern in resist and then
transfer to an underlying substrate, either
by etching or through metallization.
Fundamentally, e-beam lithography's resolution is limited
by the choice of resist, the physics of the
development process, and the subsequent
pattern transfer step. Making features smaller
in width than 10 nm (roughly 40 atoms) is
exceedingly hard, even
in isolation. Doing that regularly, at production
speeds with sub-10 nm registration across
a 30 cm wafer, is industrially unachievable
right now.
As far as I can tell, this is not a breakthrough
in any way, shape, or form. This kind of
overhype worries me. It's almost worse than
the utopian claptrap from people like Drexler -
everyone with a clue know Drexler is a loon,
but people may actually believe spokespeople
from JPL....
Slashdot Mirror
Of course, you still need to hike around a whopping big capacitor bank to have this work over any reasonable distance, and the repeat rate of fire would probably be lousy since the capacitors would have to be recharged....
First, in principle you can prepare a superposition of all possible inputs to your program. Run the program once. You've now got a superposition of all possible outputs that can be generated from your inputs.
Second, within the program itself, performing an operation some number of times N can lead to superpositions containing ~ exp(N) terms. That is, with a linear number of operations you can generate an exponentially large number of states.
You can read more about quantum computation here or here .
As part of my own research, I'm very very familiar with the Schon investigation. There are definitely questions about his PhD work. I would be willing to wager a large sum of money that the university wrote it up this way to save face, rather than admit that Schon snowed them and received a PhD for faked data.
There is no way on earth that there will be a general purpose 'fabricator' within twenty years. I'm not saying that I wouldn't like one - it's just not physically realistic!
The founders of this little group have no technical background in this field at all. Through sheer chutzpah, they've declared themselves to be authorities on nanotechnology because they've read popular treatments like those by Drexler. I'm really sick of people with no real scientific credibility becoming pundits merely by being loud.
Unless you actually understand the technology, you can't assess it reasonably! . How hard is this to figure out?!
What kind of search engine turns up no links (none, zero, zip, nada) when searching for "bell labs" ?
I assume their server is aflame or something.
Dr. Ignatiev at the University of Houston is in charge of a NASA-funded effort to develop "Space Vacuum Epitaxy", where researchers try to take advantage of the vacuum of space as a growth environment. The problem is, near-earth space simply isn't low enough pressure to work well, meaning that the researchers need to have their experiments behind a "wake-shield" - basically a big metal plate that pushes unwanted gas molecules out of the way.
The upshot is that space vacuum epitaxy is hugely expensive. Better and more controlled vacuums may be achieved on earth at a tiny fraction of the cost of a single shuttle launch. While I'm glad NASA funds research efforts like this, many such projects, even while producing cool results, often end up with the conclusion that doing the work in space is neither necessary nor desirable.
I'm a solid state physicist, too, and I've taken classes from Laughlin and Zhang, and met both Pines and Gross.
Umm, the article was quoting a Nobel laureate.
If Laughlin wants to use those words, he has the right.
Zhang started out as a high energy theorist (and a successful one at that) before getting interested in condensed matter. He's more than mathematically capable of doing GR.
Now you're just being peevish. Any CM person studying correlated electronic systems, for example, like doing transport measurements in dilute 2d electron gases in GaAs, is certainly interested in how emergent phenomena occur. Wouldn't you say that lots of CM physicists are interested in the emergence of the collective superconducting state from a (strange) metallic normal state in high-Tc superconductors? Anderson was right - more is different.
The language in the article is poorly worded here. However, your point, while strictly true, is not very useful. Come on - we could describe the motion of all the air molecules in this room using Newton's laws, but isn't the collective language of statistical mechanics (pressure, temperature, chem. potential) more useful? Your reductionist point of view is the antithesis of statistical physics that you claim to understand.
Give me a definition of "particle" that makes this self-evident. In quantum field theory, electrons are just as real as phonons. They're quantized excitations of a quantum field.
For someone with a PhD in soft CM, your viewpoints seem pretty unusual. Zhang and Laughlin are two of the smartest people around, and their ideas have a lot of appeal. Isn't it amazing that, out of 10^23 strongly interacting particles in 1 cc of copper, we still end up with excitations (quasiparticles) that act a lot like free electrons? Why are you so deeply offended by the idea that the excitations we are familiar with are actually statistical, collective properties of some other fields? Is that any more ugly than postulating the standard model, complete with its weird, CP-violating lagrangian?
Science magazine's online section right now.
That's subscription only, unfortunately. The print version will appear in Science in a few weeks.
The reason there's no preprint circulating is that Science (and Nature) are notoriously draconian about that sort of thing. Papers are embargoed until the date of publication.
A real paper does exist, though - I've got it right in front of me.
Gain is a common figure of merit of transistor-based amplifier circuits. The gain of a voltage amplifier is defined as the ratio of the size of the output signal to the size of the input signal. An amplifier that could take a 0.5V amplitude sine wave as its input and produce a 5V amplitude sine wave as its output has a gain of 10. You don't get something for nothing, of course - the amplifier has to be connected to an external power source.
A transistor is a three-terminal device. In a typical computer chip, these three terminals are called the source, the drain, and the gate. For a given voltage between the source and drain, the current that flows into the drain is strongly dependent on the voltage applied to the gate. That's what allows transistors to be used as switches: you can make a transistor that won't let current flow from source to drain unless the gate voltage is turned up past some value.
Achieving actual gain in a single-molecule device is important. Without gain greater than one, it's not possible to efficiently chain large numbers of transistors together to manipulate signals. A strong input would get degraded with each stage of transistor manipulation, eventually falling to a level too small to drive subsequent transistors.
There are *many* problems with the idea of using individual molecules to replace Si devices. Achieving a gain > 1 is a necessary but by no means sufficient step for eventual molecule-based computers. As a physicist, I think it's important to recognize real achievements in this field, but not to buy into the hype unquestioningly.
I've done extensive nanofabrication, and these guys have chosen their words so carefully as to be misleading. When they talk about making structures on the "subcellular" scale for biological research, it sounds impressive but really isn't. A typical red blood cell is 5 microns across. The smallest features produced photolithographically for your Athlon are 0.13 microns across. Even more annoying is their claim of molecular and submolecular scale device size without actually naming a number. Molecules can be big - DNA can be many microns long when uncoiled.
A meaningful figure of merit for resolution is: how small a feature can you pattern in resist and then transfer to an underlying substrate, either by etching or through metallization. Fundamentally, e-beam lithography's resolution is limited by the choice of resist, the physics of the development process, and the subsequent pattern transfer step. Making features smaller in width than 10 nm (roughly 40 atoms) is exceedingly hard, even in isolation. Doing that regularly, at production speeds with sub-10 nm registration across a 30 cm wafer, is industrially unachievable right now.
As far as I can tell, this is not a breakthrough in any way, shape, or form. This kind of overhype worries me. It's almost worse than the utopian claptrap from people like Drexler - everyone with a clue know Drexler is a loon, but people may actually believe spokespeople from JPL....