That's actually not so bad a picture, since it's about transferring and controlling angular momentum rather than charge. But the electrons do in fact leave their atoms. In this respect it's better to think of them as waves: They form waves filling out the entire material, it's just that the wave directions all cancel out until you apply a voltage. That also allows them to transfer their angular momentum to each other better, since the magnetic interaction between two electrons sitting isolated from each other is so weak that it's completely drowned out by thermal fluctuations at room temperature.
The difference is that there's very little energy required to flip a spin. The energy losses (at least the unavoidable ones) in computation come from deleting information, for instance to delete a single bit signified by a bunch of electrons caught in some place you'd let them flow to the ground line. If the voltage difference was, say, 2V you'd lose 2 eV per electron. If the same electrons had the bit stored in their spins, the energy you'd lose could be orders of magnitude lower, quite possible milli- or microeV.
Well, I'm not a particle physicist either, but I did my master's thesis on this type of system(*)
So, how close are they to applications?
First of all, they've measured from 5K up to 80K which is about quarter of room temperature, practically there by solid state physics standards (see the Nature paper). Considering that the effect didn't dwindle by more than half in that range, that's a very good sign that it could be brought up to room temperature. The problem is in getting the electrons "lined up" enough. In the Nature paper they estimate that they see about 30% more spin up than down electrons, but for real applications you'd like to get a lot closer to "100% polarization". I guess that problem might be solvable, but it includes a lot of putting very thin layers of material on single crystal with quite extreme tolerances. Then again, chip fab tolerances are quite extreme already.
In cases like this it's hard to figure out how well stuff will work without actually trying, and that's what this recent paper is about: They've built a transistor-like device with technology similar to that which would be used for mass production and measured a (tiny) effect. Now it's a matter of optimisation, and they might just get there, but it'll be years at least before it's time to start drafting the chip layout.
(*) Hint: If it involves building huge accelerators to crash particles together at pseudo-Big-Bang conditions, it's particle physics. If it involves sticking little pieces of semiconductor into a magnet at 5 Kelvin it's solid state physics. Not that solid state physicists don't use particles, we're all over electrons, phonons, excitons, magnons and many other 'ons that ignorants dismiss as "quasi"particles. Hrmph!
No, it's a quantum mechanical thing. Seeing spin as the electron spinning is a very intuitive picture, but also quite wrong. The real understanding of spin involves stuff like non-commuting operators that I won't go into here (and a quantum mechanics textbook will probably do it better) but the upshot is:
A single spin 1/2 particle like an electron can be in two states. You can measure the spin along x, y or z direction as you wish, and the answer will always come out to either plus or minus hbar/2. (hbar is the reduced Planck's constant).
It's with this spin like momentum and position: You can't know both at once. If you measure a certain spin along x, the spin in y and z directions will be in a Schrödinger's cat like state: Both + and - at once, but if you measure it you'll only see one. Of course you can choose an axis very close the original x, and you'll be very likely (but not sure) to measure the same general direction.
In sum, the easy thing to do with a spin is to treat it like a single bit, just pic one direction and measure along that. (For computer operation you'd probably like using 10-10000 spins at once to limit the effects of noise, just like transistors aren't quite reliable using a single electron (yet)). If you're into the advanced stuff you can have the spin hold one qubit (google it) and do quantum computation, but the technology in this particular report is likely to stay on the classical side.
I had a lecturer who explained that when applying for grants you'd always like the research to have imminent application. On the other hand, if you put the deadline too early you, or the people who granted the money, might have to face responsibility for the failure. In between was there was a sweet spot, which he gauged to be around 15 years or so. Ever since then I've honored him by referring to this phenomenon as the "Flensberg Optimum".
It's planned to go running next spring, but it's already at 3.5 PB. (The old LEP collider working in the same tunnel produced quite a bit of data too).
It's been replaced by ROOT (root.cern.ch). I used PAW for my bachelor's project, and I really didn't like it and I suspect you only will if you're also into FORTRAN (77!). But then again I was young and naive at that time - now I'm only naive...
And again, one may not agree with what the Party in China calls a threat to society, but they rule a sovereign nation, and it is their right to make up the minds of everyone in China about this.
There. Now I've corrected this to clarify what you were really saying.
Assuming the laser needs to be concentrated on the same spot for 1 second, the aircraft will have traveled nearly a mile.
Another thing is that the plane will have less problems getting rid of the induced heat. A sattelite can only decrease the temperature in the laser spot by conducting it to elsewhere on the sattelite, and/or radiating it into space. We can compare this to the plane which is stuck with much of the energy released in the engine, and must cool itself in the air flow.
An enemy firing a laser must either put enough power into so small a spot that it cannot be conducted to elsewhere on the plane, or he must put in a total power comparable to that of the plane's engine which is hard to do on a reflective surface.
Iran is perhaps the most extreme fundamentalist muslim nation in the world.
OTOH, don't take that to mean that Iranians are some of the most extreme fundamentalist muslims in the world - in fact many of them resent Islam for the heavy-handed version of it they've been exposed to. I recognize that the ones I've met in Europe don't represent the conservative rural population, but it's a very mixed picture. I'm actually not sure they're more religious than Americans on an individual level. With so many of them having experienced the backsides of theocracy I see plenty of hope for a democratic Iran in my lifetime.
if you really want a non-oversubscribed link, be prepared to pay $500 and up a month for it.
Just do what some ISPs in my area (Denmark) are doing: They offer 2Mbit/s very cheaply, but with download fees after the first 100MB or so. Other offers include a specified bandwidth full time, and this bandwidth is usually delivered.
That's a free market the American Way, the ISPs are safe from complaining consumers, and no one has to has to worry all day about how much they can max out their torrents without being throttled.
Re:This is a lousy solution
on
Interstellar Ark
·
· Score: 2, Insightful
I wrote an article a while back on a FAR better, obvious approach on usenet.
I think it's worse and non-obvious. A few questions:
These pebbles would enter a long ring of magnets in the spacecraft's engine, be deaccelerated to rest relative to the spacecraft with their energy stored in accumulators.
Are the pebbles charged? How do they keep their charge while moving through the solar wind? How large and strong a system of magnets/induction coils do you need to turn relativistic charged pebbles around? (Hint, bigger than "a few meters"). If they're not charged, are they magnetic? (In that case, they'll be sucked INTO the field, DEcelerating the craft). If they're neither charged nor magnetic, why do you think they'll be affected by a magnetic field?
You don't carry human crew, but self replicating machines.
Now, that's a good idea. But not a new one.
Quantum teleportation (a practical technique, demonstrated in the lab) would be used to transmit the key memory state molecules of a human brain.
And why would you use quantum teleportation for that? How do you get at the "key memory state molecules" inside a brain, and do you intend it to operate afterwards? (If you do, and if quantum state is really so necessary to "uploading", have you considered the no-cloning theorem?). And when you have transferred the state to photons, how does the transmission work across light years? (Remember, you need single photon efficiency, or someones memory will end up jumbled...)
To be blunt: You don't know what you're talking about, and neither do those who modded you up. Take a couple of physics courses.
Why can't those of us who prefer buy-to-own just download in a standard, non-defective format? Then you can put time-expire DRM on the market once you've made it work in an open and efficient fashion (or invented a free-energy machine, whichever comes first).
I would be really happy to work on a way to take the most practical 10,000 pages of the Wikipedia in a few languages and put them onto some physical media that doesn't require tech to read, and doesn't deteriorate much over time.
I hope you realize that encyclopedias (well, they're not quite Wikipedia, but still well-edited works) printed on paper are scattered in homes and libraries all around the globe already. Storing stuff on paper as a civ-collapse insurance is pretty far already, at least way further than it's been at any other point in history.
Funny isn't it, how no matter how many times humans start over with a utopian system, they end up concentrating their wealth into a small number of strong leaders and leaving a large number of impoverished citizens.
Really? That's not the tendency I see in Denmark (or most other European countries). Today wealth really is distributed somewhat evenly, at least compared to 2-300 years ago when a poor peasantry was ruled over by a bunch of nobles and an (in principle) absolute king.
Maybe that's because the development towards a democratic welfare state has happened in baby steps along with a tremendous technological change (instead of "Starting Utopia Today"). In any case, your initial assertion stands completely without documentation.
The results are presented at the IEDM conference, and it seems that there's no published article on this yet. From this page I get:
Ultra-Thin Phase-Change Bridge Memory Device Using GeSb
Y.C. Chen, C.T. Rettner***, S. Raoux***, G.W. Burr***, S.H. Chen, R.M. Shelby***, M. Salinga***, W.P. Risk***, T.D. Happ*, G.M. McClelland***, M. Breitwisch^, A. Schrott^, J.B. Philipp*, M.H. Lee, R. Cheek^, T. Nirschl**, M. Lamorey^^, C. F. Chen, E. Joseph^, S. Zaidi*, B. Yee^, H. L. Lung, R. Bergmann*, and C. Lam^, Macronix International Co. Ltd., *Qimonda, **Infineon Technologies, ***IBM Almaden Research Center, ^IBM Watson Research Center, ^^IBM Essex Junction, San Jose, CA
An ultra-thin phase-change bridge (PCB) memory cell, implemented with doped GeSb, is shown with 100microAmp RESET current. The device concept provides for simplified scaling to small cross-sectional area (60nm squared) through ultra-thin (3nm) films; the doped GeSb phase-change material offers the potential for both fast crystallization and good data retention.
Yes and no. This experiment is the first of its kind. It got published in Nature. That doesn't happen unless you've got top notch people working with state-of-the-art of equipment. Not for home tinkerers.
The big question is, can this be refined into something practical for production in a huge fab and use at home. The press release says this can be stable at high (read: room) temperature, and as far as I can tell (IAAPhysicist) this seems reasonable enough. (Refer to paramagnetic limit.) No liquid helium needed.
I foresee some problems since they must address individual bits with a combination of an oscillating field and a short RF pulse. This is hard to do, since the width of the RF beam will be much larger than the distance between bits in a useful device. Maybe a write head can be constructed making a very concentrated oscillating field.
In any case, from skimming the paper there's no blindingly obvious effects stopping it. Very sophisticated techniques were used to understand the system, but in a real device you don't care about time-resolved images of individual bits as long as you can be reasonably sure that they switch when you want them to. Lots of problems will have to be solved, but not more than usual after the first proof-of-concept of a new technology. I think these people earned their grants, both from a scientific and an economic viewpoint, and it's very easy to argue that developing it further is a good gamble.
I've had this happen to me. Not in an MRI machine, but in a liquid He-cooled cryostat. (Oxford Instruments Spectromag)
The magnet is superconducting, but the small shunt completing the circuit is heated to become a normal conductor while slowly pumping current in through some big-ass copper cables. Now in this case the copper cable connector had not been plugged in well enough, and as the field reached about 3T, the copper heated and brought the circuit out of superconductivity. The energy stored in the field was dumped within a second or two and Helium spewed out of the safety valve in a neat display of boiling air. Cost some hours of work and about 20L of (expensive) liquid He.
The cryostat is built to handle this, but an MRI machine is bigger and often has a stronger field. If it doesn't have some well-designed safety valves it's easy to imagine the Helium bath blowing out.
that doesn't account for rogue objects (those with either highly elliptical or hyperbolic orbits
If it has a hyperbolic orbit it's from outside the solar system. Not from "beyond Plu^W^W^W Neptune" but from interstellar space and on its way out again. AFAIK we haven't spotted any such objects yet (apart from cosmic radiation) and if we do, I'd almost hope it impacts just so we can get our hands on the samples.
But note that the MR2A16A is only 4Mbit. NEC has a 16 Mbit MRAM out, but you'll get nowhere the same capacity as with flash for your space and money.
I'm kinda partial to MRAM since I do my thesis in spintronics, but it's just not there yet. Maybe some MRAM-buffered flash system would be handy, but that goes a bit beyond doing your own circuit boards.
Slashdot Mirror
That's actually not so bad a picture, since it's about transferring and controlling angular momentum rather than charge. But the electrons do in fact leave their atoms. In this respect it's better to think of them as waves: They form waves filling out the entire material, it's just that the wave directions all cancel out until you apply a voltage. That also allows them to transfer their angular momentum to each other better, since the magnetic interaction between two electrons sitting isolated from each other is so weak that it's completely drowned out by thermal fluctuations at room temperature.
The difference is that there's very little energy required to flip a spin. The energy losses (at least the unavoidable ones) in computation come from deleting information, for instance to delete a single bit signified by a bunch of electrons caught in some place you'd let them flow to the ground line. If the voltage difference was, say, 2V you'd lose 2 eV per electron. If the same electrons had the bit stored in their spins, the energy you'd lose could be orders of magnitude lower, quite possible milli- or microeV.
Well, I'm not a particle physicist either, but I did my master's thesis on this type of system(*) So, how close are they to applications?
First of all, they've measured from 5K up to 80K which is about quarter of room temperature, practically there by solid state physics standards (see the Nature paper). Considering that the effect didn't dwindle by more than half in that range, that's a very good sign that it could be brought up to room temperature. The problem is in getting the electrons "lined up" enough. In the Nature paper they estimate that they see about 30% more spin up than down electrons, but for real applications you'd like to get a lot closer to "100% polarization". I guess that problem might be solvable, but it includes a lot of putting very thin layers of material on single crystal with quite extreme tolerances. Then again, chip fab tolerances are quite extreme already.
In cases like this it's hard to figure out how well stuff will work without actually trying, and that's what this recent paper is about: They've built a transistor-like device with technology similar to that which would be used for mass production and measured a (tiny) effect. Now it's a matter of optimisation, and they might just get there, but it'll be years at least before it's time to start drafting the chip layout.
(*) Hint: If it involves building huge accelerators to crash particles together at pseudo-Big-Bang conditions, it's particle physics. If it involves sticking little pieces of semiconductor into a magnet at 5 Kelvin it's solid state physics. Not that solid state physicists don't use particles, we're all over electrons, phonons, excitons, magnons and many other 'ons that ignorants dismiss as "quasi"particles. Hrmph!
No, it's a quantum mechanical thing. Seeing spin as the electron spinning is a very intuitive picture, but also quite wrong. The real understanding of spin involves stuff like non-commuting operators that I won't go into here (and a quantum mechanics textbook will probably do it better) but the upshot is:
A single spin 1/2 particle like an electron can be in two states. You can measure the spin along x, y or z direction as you wish, and the answer will always come out to either plus or minus hbar/2. (hbar is the reduced Planck's constant).
It's with this spin like momentum and position: You can't know both at once. If you measure a certain spin along x, the spin in y and z directions will be in a Schrödinger's cat like state: Both + and - at once, but if you measure it you'll only see one. Of course you can choose an axis very close the original x, and you'll be very likely (but not sure) to measure the same general direction.
In sum, the easy thing to do with a spin is to treat it like a single bit, just pic one direction and measure along that. (For computer operation you'd probably like using 10-10000 spins at once to limit the effects of noise, just like transistors aren't quite reliable using a single electron (yet)). If you're into the advanced stuff you can have the spin hold one qubit (google it) and do quantum computation, but the technology in this particular report is likely to stay on the classical side.
I had a lecturer who explained that when applying for grants you'd always like the research to have imminent application. On the other hand, if you put the deadline too early you, or the people who granted the money, might have to face responsibility for the failure. In between was there was a sweet spot, which he gauged to be around 15 years or so. Ever since then I've honored him by referring to this phenomenon as the "Flensberg Optimum".
At least according to this article.
It's planned to go running next spring, but it's already at 3.5 PB. (The old LEP collider working in the same tunnel produced quite a bit of data too).
It's been replaced by ROOT (root.cern.ch). I used PAW for my bachelor's project, and I really didn't like it and I suspect you only will if you're also into FORTRAN (77!). But then again I was young and naive at that time - now I'm only naive...
An enemy firing a laser must either put enough power into so small a spot that it cannot be conducted to elsewhere on the plane, or he must put in a total power comparable to that of the plane's engine which is hard to do on a reflective surface.
That's a free market the American Way, the ISPs are safe from complaining consumers, and no one has to has to worry all day about how much they can max out their torrents without being throttled.
I wrote an article a while back on a FAR better, obvious approach on usenet.
I think it's worse and non-obvious. A few questions:
These pebbles would enter a long ring of magnets in the spacecraft's engine, be deaccelerated to rest relative to the spacecraft with their energy stored in accumulators.
Are the pebbles charged? How do they keep their charge while moving through the solar wind? How large and strong a system of magnets/induction coils do you need to turn relativistic charged pebbles around? (Hint, bigger than "a few meters"). If they're not charged, are they magnetic? (In that case, they'll be sucked INTO the field, DEcelerating the craft). If they're neither charged nor magnetic, why do you think they'll be affected by a magnetic field?
You don't carry human crew, but self replicating machines.
Now, that's a good idea. But not a new one.
Quantum teleportation (a practical technique, demonstrated in the lab) would be used to transmit the key memory state molecules of a human brain.
And why would you use quantum teleportation for that? How do you get at the "key memory state molecules" inside a brain, and do you intend it to operate afterwards? (If you do, and if quantum state is really so necessary to "uploading", have you considered the no-cloning theorem?). And when you have transferred the state to photons, how does the transmission work across light years? (Remember, you need single photon efficiency, or someones memory will end up jumbled...)
To be blunt: You don't know what you're talking about, and neither do those who modded you up. Take a couple of physics courses.
Why can't those of us who prefer buy-to-own just download in a standard, non-defective format? Then you can put time-expire DRM on the market once you've made it work in an open and efficient fashion (or invented a free-energy machine, whichever comes first).
I would be really happy to work on a way to take the most practical 10,000 pages of the Wikipedia in a few languages and put them onto some physical media that doesn't require tech to read, and doesn't deteriorate much over time.
I hope you realize that encyclopedias (well, they're not quite Wikipedia, but still well-edited works) printed on paper are scattered in homes and libraries all around the globe already. Storing stuff on paper as a civ-collapse insurance is pretty far already, at least way further than it's been at any other point in history.
Funny isn't it, how no matter how many times humans start over with a utopian system, they end up concentrating their wealth into a small number of strong leaders and leaving a large number of impoverished citizens.
Really? That's not the tendency I see in Denmark (or most other European countries). Today wealth really is distributed somewhat evenly, at least compared to 2-300 years ago when a poor peasantry was ruled over by a bunch of nobles and an (in principle) absolute king.
Maybe that's because the development towards a democratic welfare state has happened in baby steps along with a tremendous technological change (instead of "Starting Utopia Today"). In any case, your initial assertion stands completely without documentation.
The results are presented at the IEDM conference, and it seems that there's no published article on this yet. From this page I get:
Ultra-Thin Phase-Change Bridge Memory Device Using GeSb
Y.C. Chen, C.T. Rettner***, S. Raoux***, G.W. Burr***, S.H. Chen, R.M. Shelby***, M. Salinga***, W.P. Risk***, T.D. Happ*, G.M. McClelland***, M. Breitwisch^, A. Schrott^, J.B. Philipp*, M.H. Lee, R. Cheek^, T. Nirschl**, M. Lamorey^^, C. F. Chen, E. Joseph^, S. Zaidi*, B. Yee^, H. L. Lung, R. Bergmann*, and C. Lam^, Macronix International Co. Ltd., *Qimonda, **Infineon Technologies, ***IBM Almaden Research Center, ^IBM Watson Research Center, ^^IBM Essex Junction, San Jose, CA
An ultra-thin phase-change bridge (PCB) memory cell, implemented with doped GeSb, is shown with 100microAmp RESET current. The device concept provides for simplified scaling to small cross-sectional area (60nm squared) through ultra-thin (3nm) films; the doped GeSb phase-change material offers the potential for both fast crystallization and good data retention.
Yes and no. This experiment is the first of its kind. It got published in Nature. That doesn't happen unless you've got top notch people working with state-of-the-art of equipment. Not for home tinkerers.
The big question is, can this be refined into something practical for production in a huge fab and use at home. The press release says this can be stable at high (read: room) temperature, and as far as I can tell (IAAPhysicist) this seems reasonable enough. (Refer to paramagnetic limit.) No liquid helium needed.
I foresee some problems since they must address individual bits with a combination of an oscillating field and a short RF pulse. This is hard to do, since the width of the RF beam will be much larger than the distance between bits in a useful device. Maybe a write head can be constructed making a very concentrated oscillating field.
In any case, from skimming the paper there's no blindingly obvious effects stopping it. Very sophisticated techniques were used to understand the system, but in a real device you don't care about time-resolved images of individual bits as long as you can be reasonably sure that they switch when you want them to. Lots of problems will have to be solved, but not more than usual after the first proof-of-concept of a new technology. I think these people earned their grants, both from a scientific and an economic viewpoint, and it's very easy to argue that developing it further is a good gamble.
History disagrees with you.
I've had this happen to me. Not in an MRI machine, but in a liquid He-cooled cryostat. (Oxford Instruments Spectromag)
The magnet is superconducting, but the small shunt completing the circuit is heated to become a normal conductor while slowly pumping current in through some big-ass copper cables. Now in this case the copper cable connector had not been plugged in well enough, and as the field reached about 3T, the copper heated and brought the circuit out of superconductivity. The energy stored in the field was dumped within a second or two and Helium spewed out of the safety valve in a neat display of boiling air. Cost some hours of work and about 20L of (expensive) liquid He.
The cryostat is built to handle this, but an MRI machine is bigger and often has a stronger field. If it doesn't have some well-designed safety valves it's easy to imagine the Helium bath blowing out.
that doesn't account for rogue objects (those with either highly elliptical or hyperbolic orbits
If it has a hyperbolic orbit it's from outside the solar system. Not from "beyond Plu^W^W^W Neptune" but from interstellar space and on its way out again. AFAIK we haven't spotted any such objects yet (apart from cosmic radiation) and if we do, I'd almost hope it impacts just so we can get our hands on the samples.
But note that the MR2A16A is only 4Mbit. NEC has a 16 Mbit MRAM out, but you'll get nowhere the same capacity as with flash for your space and money.
I'm kinda partial to MRAM since I do my thesis in spintronics, but it's just not there yet. Maybe some MRAM-buffered flash system would be handy, but that goes a bit beyond doing your own circuit boards.
I'm sure it's possible to combine several 128K lines to get one single hi-speed line.
Suppose you and 5 or 6 of your neighbours had 128K each. How would you go about it?
Another parallel: The EU has the authority to regulate basically anything regarding the internal market. Look up "interstate commerce".
In Denmark the Conservative and the Liberal Party are both right wing parties. They currently form the government together.
The name of the liberal party is "Venstre" meaning "left". Go figure.