Slashdot Mirror
IBM Tech Detects & Changes Spin of Single Electron
An anonymous reader writes "Looks like we have another step forward in Quantum Computing - IBM has discovered how to detect and change the spin of a single electron. Won't be long before we're all solving impossible encryption problems.
"
So this allows read-write of qubits, right?
My Photography - http://ian-x.com
The Deathlings (comic) - http://thedeathlings.com
So what do we do if quantum computers can decrypt anything in almost real-time?
All I can think of is making the data streams uninterceptable, which leads us back to encoders/decoders built using quantum entanglement.
**TODO** Steal someone elses sig.
A while back there was a proposal to have a public onetime pad system that worked like this. there is a server, perhaps a sattelite, that is streaming random numbers at say gigabytes per second. To encode a message you weakly encrypt a prior message to the recipient telling him a precise start time: say the message reads: start colleting your onetime pad at the first occurence of the first 5 digits of the number pi that come after 12 noon. you both then collect the data that comes at that time and treat ti as a shared one time pad.
you opponents may be able to decrypt the pre-message eventually but not it time to make the start time. thus they cant collect the onetime pad data. the data rate of the random stream is chosen so that no plausible storage system could retain more than say a few hours worth of the data, so no one could just record it all. As long as no one can crack your message on that time scale you can dsafely send the one time pad whihc no one can crack by technical means.
Some drink at the fountain of knowledge. Others just gargle.
Achille Talon
Hop!
Mix the principle from this Random Lava Lamp Generator, a high speed multi-mega pixel digital and film with a suffiviently high resolution, pipe resulting image as raw key-data, and you're done...
or if you prefer "better randomness" (sic) use HotBits cesium decay generated random numbers and pipe...maybe not gigabytes of data ( (about 30 bytes per second, to be precise, sucks...but then...
Mix both obtained key with the obscure, non repetitive algorithm of your choice (a simple XOR will be enough) and you can start having pretty impressive figures...
Generating gigabytes of data is quite easy, an I agree with you that making sure they are truly random is much more difficult, but it's not really that impossible 8)
The real problem is making sure both people get the SAME timepad, the problem here being to have both people REALLY in synch when startng the capture to generate the pad...
which is another problem entirely...
It takes 40+ muscles to frown, but only four to extend your arm and bitchslap the motherfucker
You know, I've never seen an explanation of why they went to copper interconnects and not silver. Silver is a better conductor, and the material cost doesn't seem significant.
While I don't KNOW, I can speculate:
1) Silver oxidizes (tarnishes) very easily. Silver oxide is NOT such a great conductor. Given the extremely small size of the interconnects, this could lead to chips with EXTREMELY short service lives - like minutes. (Aluminum - copper's predecessor - oxidizes, too, but the oxide has about the same atomic spacing as the metal and is very hard {saphire}, producing a protective armor that stops further oxidization. Copper oxidizes very slowly unless strongly heated.)
2) Even if they could somehow passivate the silver conductors, there's still the question of whether the silver atoms would "electromigrate" into the chip, driven by the electric fields. This is why gold metalazation (initially believed to be nearly ideal due to the difficulty of oxidizing gold) was abandoned near the start of the integrated circuit era, after the effect was discovered when the "purple plague" wiped out most of the high-tech devices - mostly discrete components in military electronics - that used it.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Step 1. Get a 1-bit quantum computer working
Step 2. Wire 1,024 of them together
Step 3. Break 1,024-bit encryption!
In reality, you now have the capability to solve 1,024 separate 1-bit problems. To solve a 1,024-bit problem, the electrons carrying each qubit need to be entangled with each other.
Keeping things in a state of quantum entanglement is extremely difficult. The most I've ever read about was 7 qubits entangled for less than a microsecond. Note that as the number of entangled objects (particles or molecules) increases, the operation gets exponentially harder. As the time to complete an operation increases, it also gets exponentially harder. Quantum computing boosters won't tell you this, but it is not just a matter of getting a prototype working and then making it bigger.
This is why gold metalazation (initially believed to be nearly ideal due to the difficulty of oxidizing gold) was abandoned near the start of the integrated circuit era, after the effect was discovered when the "purple plague" wiped out most of the high-tech devices - mostly discrete components in military electronics - that used it.
Just looke it up, and it appears I've been confused about the purple plague. Actually it's not electromigration of gold into silicon, but an effect of ultrasonically bonding gold wires to aluminum metalization. This forms several alloys between the gold and the aluminum (one of them purple) and their differential diffusion leads to voids and eventuall failure of the bond.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Random points:
The IBM research has taken over from bell labs as being one of the best research labs around. It is such a shame bell labs went from being amazing to depressing but that is a different story.
Doubly so since it didn't have to happen.
Bell Labs was originally intended to be a boondoggle, since part of the deal for Bell's original monopoly was that they could set phone rates to make a particular profit (6%?) on every dollar they spent on building the phone system, including research. So they set up a R&D arm that was mandated to spend as much money as possible on research with some vuage connection to telephony, in the expectation of being able to make 6 cents on every dollar spent.
It was a "failure" from the first year: They were PROFITABLE, earning/saving more money from using and licensing the results of the research than they spent on doing it. On the average, basic research pays off big-time (even if you can't tell in advance what any particular project might produce). Example: The transistor.
Unfortunately, after the dissolution of the monopoly, the successor to Bell Labs became infected with the "Harvard Business School" style of short-term milking: Cut R&D (which costs money now and pays off later), creating a temporary boost in the profit figures followed by a collapse. Declare that you're a genius, cash out and move on to the next sucker company, leaving your successors to take the blame when the house of cards collapses.
Fortunately, IBM has learned both of these lessons of Bell Labs, big-time (as well as the Labs' UNIX lesson of how {essentially} giving away source code leads to lots of business for computer companies) and has become a worthy successor.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
This could lead to incredibly high storage density [cat seek time joke deleted]"
But seriously: Did you notice some of the other related experiments from the same group, listed at the bottom?
One struck me: The inclusion of a single ferromagnetic impurity atom in a semiconductor, with its magnetization state producing a srong and extremely localized effect on the electronic properties of the semiconductor.
This might lead to a RAM where the storage element is a transistor with a single magnetic atom embedded in the "gate" region, turning it on or off depending on the spin of the single magnetizing electron.
Extremely tiny. Extremely fast to read (probably a ballistic-transport FET). Extremely fast to write (electron spin flips REALLY fast). Extremely low power (1/2000 electron volt needed to flip the spin).
And that's not even the most impressive thing in the list.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
http://news.nanoapex.com/
;-)
Why won't just ask Slashdot to post this link as permanent (along with wired.com )? I read it every day. They have diff. sections: nanotubes, nanoelectronic, nanoenergy, MEMS (nanomachines),... the first time I looked at it i became a nanoaddict
Just look at these nanoapex news regarding spintronics (reverse sorted by date):
Sep 03: Spintronics Breakthrough: Negative Resistance of a Single Magnetic Domain Wall Measured
Jun 23: Physicists Build New Microscope to Study Electron Spin
Apr 26: IBM, Stanford Collaborate on World-Class Spintronics Research
Mar 22: Silicon-based magnets boost spintronics
Feb 28: Spin valves open organic chip era
And that was just ONE topic. 'nuff said.
No, it will not work. Assume, my pre-message is use NBC + ESPN as the feed. The pre-message is decrypted after 1 week. I assume the evesdropper can get the archives form the respective stations and decode the message.
Same principle applies to using a paticular satellite. We can get information from that broadcast station and replicate the feed. Agreed, all this is not possible for the common man. But then RSA with higher bits is sufficient for the common man.
That was the reason I suggested we use some natural source like cosmic background radiation or some other random event in space which everyone can observe. No one has the abilty to record all the events simultaneously for long time (at least currently and in the near future). But it also makes the system very less practical.
Python script to convert photos into "artsy" portraits: http://p2pbridge.sf.net/pyPortrait/
How exactly does quantum computing allow us to solve impossible encryption programs? I may not be a quantum physicist, but I do have some basic understanding. Of course I have a basic understanding of electronics as well but no idea how to make an AND gate out of transistors....
The bigger use I see for this new technology isn't encryption, it is end-to-end communication without the possiblity of interception. In this event, you simply don't need to encrypt the transmission because it is impossible to intercept it anywhere except the transmitting and receiving devices.
Also could have great implications for the space program as such information "travels" faster than the speed of light. Travels is in quotes because perhaps the information doesn't actually travel as we think of it.
LedgerSMB: Open source Accounting/ERP