Last updated: 2002-02-25
Here is some stuff that's tangentially related to Quantum Computing. Just something I (David Cary) threw together that I thought you might find interesting. If you find any more related resources, I would appreciate you telling me about them so I can add (links to) them to this page. Comments?
index:
[FIXME: indicate which links are introductory material, and which ones assume you already know all that stuff.]
David also maintains related files:
tutorials, kinds of quantum computers, general reviews of multiple kinds of quantum computers.
Some related Usenet Newsgroups:
From: Brad Taylor <blt at emf.net> Newsgroups: sci.nanotech Subject: Re: Nano logic elements? Date: 10 Jun 1996 11:23:42 -0400 Anders Sandberg wrote: > What about those nifty quantum cellular automata based on electrons > trapped in quantum dots I mentioned a long time ago? > o= quantum dot *= dot with electron > o * o * o * o * o * > o o o o o ... > * o * o * o * o * o There is a home page for coupled quantum dots at UA. http://cs.ua.edu/graduate/lusth/qca/index.html They are called QDCAs for Quantum Dot Coupled Arrays. As far as I can tell (if they really work), QDCAs offer an unbeatable combination of simplicity, speed and low power for computing. -- Since they are a regular array they might be manufactured by fabrication of a unit cell followed by crystallization. - Unlike "rod-logic", the program in such a crystal would be downloadable at run time. Rod logic would require the program to be known at manufacturing time. - All that is required of the unit cell is to trap a few electrons, and space them properly. This can be accomplished via a wide variety of structures, giving QDCAs a good chance for realization. - The computation is reversable and therefore low power in nature. - Brad [According to the presentation I saw a couple of years ago at PhysComp, you have to draw a circuit, so the array isn't regular and the function isn't downloadable, e.g. to make an inverter: o * o * o * o * o o o o * o * o * o * o * ... o o o * o * o * o There was also some concern among the audience about effects like signal reflection, since there isn't any signal restoration mechanism in the circuit. --JoSH]
[low power]
Date: Sat, 1 Jun 1996 00:00:04 -0400 (EDT) From: transhuman at umich.edu Subject: >H Digest ... **************** From: Eugene Leitl <Eugene.Leitl at lrz.uni-muenchen.de> Subject: >H molCAM/quantum dot array Transhuman Mailing List A Short Discussion of Concepts Relevant to the Molecular Cellular Automata Machine (molCAM), a Close Cousin to the Quantum Dot Array Computer ( Caveat: this is an informal fact assembly for a wide audience. Both technobabble and necessary rigor are largely missing. This speculative microessay was not meant to hold up to scienfic standards. No refs are given. ) 1. Impacts of Strong (Drexlerian) Nanotechnology Though Drexler is a major contributor to the still pretty young field of nanotechnology, he has by no means invented the infinitesimal realm of nano. Prior to the advent of his diamondoid SISD generic nanoassembler design, diverse ideas about both top-down downscaling of solid state devices and using autoorganizing and autoassembling intrinsic properties of solvated biopolymers in a bottom-up approach were already fairly widespread. The crux of Drexler's design outline is high matter throughput/ processing capability which at the same time features extremely precise, atomic, level of control. Should the design prove itself implementable, the implications are literally unimaginable. Especially, greatly facilitated creation of noticeably superrealtime, significantly above human intelligence level information processing systems are thought to cause a positive design autofeedback loop, almost instantly precipitating an extremely rapid (days to months scale) intellectual runaway, the so-called Singularity (Vernor Vinge). The Singularity Theory, or, better, conjecture, is further fueled by the well-known fact that many current development trends can be fitted to exponential or even hyperbolic functions with negligeable deviations. Extrapolation of these trends into the near future (25-40 years) thus forecasts obviously impossible, since absolutely ridiculous values. Either, the predictions are entirely wrong, and we will experience a saturation at a high level, the neo-"Golden Age", or, more likely, a catastrophic breakdown, not unlike the autoinhibition of bacterial growth in a finite medium, orelse we will enter the brave new era of trans- or posthumanism, which significance can only be compared with the primeval life nucleation event on Earth. While this may sound like pompous garbage at best, or, obscure, pseudoreligious faith at worst, as future development trajectories are thought to be fundamentally uncomputable, this view can be veriefied easily enough, as most of us will live to see it. The only fact which so far seems fairly certain, is that "the world will turn strange, soon". In fact, the continuously increasing rate of change, the first forequake of Singularity, has produced a kind of world in which native neolithical societies, though rapidly vanishing, can transiently coexist with maturing aerospace, infoprocessing and biotechnology industries. As in analytical chromatography, the wandering spots on the substate plate are being increasingly spread over a larger area, some of them almost keeping up with the solvent (shockwave) front. I dunno, "strange" may already seem a quite apt epithet. Relax, you ain't seen nothing yet. 2. Alternatives Anyway, should one strive "merely" for building a drastically better computer, without assuming imminent availability of all-purpose Drexlerian nanoassemblers, which bootstrap or even implementability window might well be too narrow for practicability, one is left with the weak, "wet" autoassembly biopolymer nanotechnology. Should we succeed in its implementation, the resulting hardware will at the very least be extremely instrumental in the subsequent bootstrap of Drexler-flavoured systems. Why biopolymers? Can't we simply utilize the numerous structures and powerful methods organic chemistry can offer us? There are two basic reasons. Contrary to the widespread scientific legend, biopolymers are by no means inefficient or flabby. Proteins are all-purpose linear polymers, nevertheless their bandwidth reaches from thixotrophic, highly hydrated gels and phospholipid-protein conglomerates to the pretty dry, noticeably better-than-kevlar-calibre cobwebs and other tough fibrous structural proteins. Thus, circuitry matrix needs not to be flabby nor heavily hydrated. Proteins have been subjected to heavy optimization during GYrs of evolution, their precipitated 20 amino acid minimum alphabet surely sufficiently all-purpose. The best reason, of course, is our already extremely large and exponentially expanding fount of knowledge about protein structure, their folding kinetics and the rapidly growing skill of manipulating (by recombinant DNA technology) the so far the only working instance of nanotechnology we know of: Biological Life. In contrast to that, the level of complexity achievable by organic chemistry is very limited. Catalytically active, organic switchable enzyme analogons bear great promise, but are terribly difficult to design and to optimize. Alas, awareness of their potential feasibility and necessity of a drastically different approach in their synthesis management has not yet become widespread in the mainstream chemical community. Further progresses in organic synthesis are only possible by further migration towards mechanosynthesis. Mechanically highly constrained, aligned systems, e.g. organic photochemistry in crystal lattice as opposed to photochemistry in (inert) solvate, show drastical deviations in their products as many configuration space trajectories are excluded by constraints imposed by the embedding crystal lattice matrix. Biochemistry already utilizes mechanosynthesis at a noticeable degree, at least at the degree possible in a nonstiff, solvated system. It should be noted, that natural biological systems rarely exploit the maximum possible stiffness of enzymes. Even preliminary results of site-directed mutagenesis optimization of antibodies has shown that natural proteins can be made more stable, sometimes even drastically so by the mutation of one to few amino acids. Another key factor is disulfide bridge crosslink. Nature cannot tolerate deposition of heavily crosslinked albeit signficiantly stiffer systems with the cell, as their creation is a highly irreversible process, which might also generate significant amounts of damaging radicals. Would this path be taken, it would lead to wastage of material, and thus of metabolic energy necessary for the biosynthesis of the monomer precursors, synthesis and subsequent secretion. Moreover, stiff, nondegradable systems are fundamentally incompatible with the homeostasis maintanance control paradigm of the biological cell. Enzymatic activity is controlled both by the total number of enzymes within the cell as their individual modification, e.g. by phosphorylation or interaction with an allosteric cofactor. This is impossible, or, at least very hard to do on a crosslinked enzyme. Further difficulty in crosslinking by disulfide bridges is, apart from the fact that there is an upper limit the the fraction of crosslinks per volume, as (a potentially large) percentile of the sequence string is necessary for the encoding of a robust folding trajectory as well as proper alignment of the complementary cystine residues. Even tiny deviation from these constraints will result in nonviable proteins. Nevertheless, the design of such systems appears possible. The biopolymers as implementation tools impose a large number of constraints, which rule out lots of possible design targets. For one, though proteins are great for autoassembly and support matrix, they are virtually useless for computation. Hence, one is forced to use the (yet almost entirely hypothetical) molecular circuitry, which works by utilizing electronically excited states of small organical and metallorganical molecules, polyenes, columnar complexes, dyes, etc. Lasers or a potential gradient can be utilized to pump a quantum system into a higher state, using several, switchable relaxation pathways for computation. The achievable switching rates are noticeably above those of diamond rod logic, power dissipation should also be smaller (?), perhaps significantly so. The upper limit of switching rate is obviously given by photodissotiation of the molecular circuitry. However, protein matrix constituting the circuitry cavity, if properly engineered, can both influence excited state lifetime and contribute to drastically enhanced robustness by providing well defined relaxation pathways for dangerous states. The need for molecular circuitry, which has to be generated by the very limited means of organical chemistry and biochemistry, obviously imposes an upper limit for atomic subsystem complexity. Instead of seeing this as a minus, one might as well see it as a challenge turning to the most minimalistic computer design known: cellular automata machines, abbreviated as CAMs. Apart from their intrinsic simplicity, CA show an excellent congruency of with nature's ways of doing computation: namely, by means of a very large number of asynchronous, purely locally coupled simple subsystems. Contrarily to common belief, most real-world problems are massively parallel. Their apparent sequentiality is merely an artefact of limited introspection; a severely biased perception of the physical reality. Furthermore, CA can greatly profit from results of complexity science, utilizing new knowledge gained in emergent behaviour studies, making adaptive, robust self-healing systems more than just a dream. Once the decision for a protein-matrix, molecular-circuitry 3d molecular automaton machine (molCAM) has been made, many things start making remarkable sense. Many proteins and large protein complexes, e.g. viruses, can be crystallized in macroscopic, optically clear crystals, which inherently shield damaging shortwave radiation. While many of the crystals are very weak mechanically (you can poke with your finger through them), some being thixotropic gels, one should not forget the fact that biopolymer crystal precipitation within the cells is greatly inhibited by the artefacting generating process, the Darwinian evolution (target has 'evolution' written in huge letters all over it). In fact, it is a miracle most proteins require only very little coaxing to crystallize at all. An engineered or de novo designed protein crystal can well form a clear, tough matrix with the circuitry embedded. Again, it might be instrumental to remember that spider silk protein significantly surpasses kevlar in its mechanical properties. Crystal growth is an emergent, self-organizing parallel process without a central organizing authority. Since the elementary cells of molCAM are so large and form intricate complementary-surface docking sites, the number of defects can be made virtually nil. Due to extensive crosslinking, both within the elementary cell and at the intercell level (complementary surface interaction play also a major part), the crystal can be made mechanically tough, resisting translocation defects. Due to stiffness and defect considerations, the connectivity must be low. Should we settle for the cubic primitive lattice (which might not be the optimum but is pretty convenient to reason about), only the six nearest neighbours can be directly contacted. Rule complexity, especially in a lookup implementation, can be greatly reduced by utilizing a rotation invariant rule. Again, due to low achievable complexity, the cell's state space must be limited to few-bit values. The protein matrix and the scaffold (auxiliary proteins that are needed transiently and are not incorporated into the final matrix) play the following two roles. First, they envelop organically or biochemically synthesized circuitry subblocks, creating complexes with differing properties of both the uncomplexed protein and the piece of circuitry. This may utilized for efficient purification (e.g. gel electrophoresis, etc.), as affinity chromatography, the utilization of extremely high specifity _and_ high binding constant of immobilized antibodies raised against certain haptenes (small antigens, usually presented on a spacer) for an excellent purification. Second, they envelop, align and join pieces of the circuitry machinery in a given sequence, until in an hierarchic assembly few-step process the individual CA cell has been constructed. Since the assembly is SIMD in one pot, albeit with interim purification, the absolute number of elementary cells to be later assembled into a CAM macrocrystal is large. The elementary cell is roughly 0.1-1.0 um sized, which may seem fairly large even by today's standards of photolithographically attainable structures. It should not be underestimated however, as the cell's complexity is noticeably higher (it is a tiny few-bit computer) than the according semiconductor structure, the cells are aligned in a true 3-lattice instead of 2-lattice of photolithography, the concentration of circuitry per volume is limited due to inherent structural overhead neccessary both for the encoding of a robust folding pathway as well as a well-defined target fold, the need of optical transparency for optical power and I/O as well as maximum tolerable power/volume dissipation. The synchroniziation problem is not entirely trivial. The application of the local composite discrete Hamiltonian upon the neighbourhood (base of the light cone) must be perfectly parallel, allowing no runtime differences to accumulate during subsequent iterations, even at utilized ns and noteceably sub-ns cycle times. As infinitely precise timers within each cell are obviously impossible, a kind of external clocking must applied. Two basic flavours are possible: clocking by laser pulses (which may be also utilized for power source) and an active clock by propagating wave phenomena. The former utilizes the high speed of light which significantly surpasses signal travel and switching time in a CAM, the latter achieves the synchronization by subsequently traveling wavefronts which are originated by a few cells, which may be random. Both assume that at a single iteration time, the infinitesimal cell differences are yet invisible. The autopropagating wave state is assumed to be immediately forwarded to the neighbours, while initiating a computation cycle after a short delay. After the computation cycle the cell is temporarily insusceptible to subsequent waves, suffering the nonexcitable 'refractory time' before restoring the original state. The Gentle Reader may grant me that the above rough hardware outline (there is more to it than mentioned here) seems to be both implementable and operable, but what about the thorny issues of interfacing and programming? [...] ...to be continued... 'gene **************************************************************
Very short, but impressive web site -- http://www.ti.com/research/docs/nano.htm [Alas, this link now seems dead.] photograph of a *real* quantum-electronic device. Texas Instruments Demonstrates Quantum Effect Transistor http://www.ti.com/corp/docs/history/quantum.htm Texas Instruments Demonstrates World’s First Quantum IC that Operates at Room Temperature http://www.ti.com/corp/docs/history/quantumic.htm
-- "Epilogue: Computing the Future", p. 194, of the book _the Quantum Dot: A Journey into the Future of Microelectronics_ book by Richard Turton 1995Starting in 1968, IBM launched a massive project to convert the fledgling Josephson junction technology into a viable form suitable for constructing a computer based on superconducting devices. Although several devices based on the Josephson effect had been demonstrated in the laboratory, the task ... required that high-density integrated circuits be mass produced at low cost using what were essentially unknown materials. After 15 years and an estimated cost of $300 billion the project came to an end. Although enormous progress had been made in this time, one of the principal factors affecting the decision to terminate the project was that in the intervening period silicon technology had advanced to such a stage that the difference in predicted performance of the two systems was down to a factor of only two.
http://vesta.physics.ucla.edu/~smolin (Quantum Information Page)
http://www.thehub.com.au/~mitch/q-comp.html (quantum computing)
Quantum Computation Reference Collection: http://feynman.stanford.edu/qcomp/
Quantum Information at Los Alamos National Laboratory http://p23.lanl.gov/Quantum/quantum.html
Date: Mon, 17 Feb 1997 12:54:52 -0800 From: Brad Taylor <blt at emf.net> Organization: Giga Operations To: David Cary Subject: Re: Quantum Computing page David Cary wrote: > > Hi everyone. I am interested in understanding quantum computing, but I > found it difficult to find any information on the subject. > > >From postings I read on Usenet News, I sensed that many of you were just as > frustrated as I was. > > So I built a crude Quantum Computing page at > http://www.rdrop.com/~cary/html/Quantum_C_FAQ.html > > Any comments ? Should I post an announcement to Usenet News ? Which newsgroups ? > > David Cary "mailto:d.cary@ieee.org" "http://www.rdrop.com/~cary" > Future Tech, quantum computing, digital hologram, PCMCIA, <*> O- Hi David: I appreciate your post. When I got access to the web I became somewhat obsessed with 'future computing' whatever that means. I think there is actually quite a bit available, but most of it is not very meaty. The range of computational devices that are acessable at the nanoscale is extremely diverse. The best stuff is in the academic archives at http://xxx.lanl.gov/ but these are very rough reading and in an obscure version of pdf/ps If you are interested in following my web trails my personal bookmarks are on line at: http://www.infowest.com/a/atay/blt/blt.htm The most relevant pages are under computonium/sources and computonium/mesoscale. A somewhat unrelated site but very interesting site is: http://pineapple.apmaths.uwo.ca/~gno/ I have been meaning to put together a page for my favorite papers. If I do I will send you a note. - Brad
"Error-Correcting Code Keeps Quantum Computers on Track" article by Barry Cipra in _Science_ 1996 April 12.
refers to "in a paper published last year in _Physical Review A_, mathematician Peter Shor of AT&T Bell Labs (now AT&T Research) has shown how -- at least in theory -- to nudge a quantum system back into line without looking at it directly. Shor's theoretical feat is now triggering a flurry of error-correction schemes based on his method."
"Seth Lloyd [of MIT], a quantum-computing pioneer"
"Before Shor came up with this idea, nobody thought it [quantum error-correcting codes] was possible."
(I've seen the article in _Scientific American_ "The Square Root of Not" -- I *think* I understand it, but the particular gate described appears pretty useless, and I don't know enough to extrapolate to more useful (multiple-input) gates).
What I'd find most interesting would be a list of proposed quantum logic gates, with a functional description of each (rather than a "how it works" description).
For example, a *functional* description of a particular low-pass filter would be "passes all frequencies from DC to F1 with a 20 dB gain with a ripple of 0.1%; blocks all frequencies from F2 on up with an attenuation of at least 80 dB." (The "how it works" description would go into the particular op amps and/or transistors used -- but since there's *many* different possible circuits that all act functionally the same, I don't find that level of detail particularly useful).
(2) Would you like me to mail you a summary of the information I've gotten from you and other sources ?
(3) Would you mind me making public any info you give me (included in the summary)? (Some people are *so* sensitive about the privacy of their email messages, that I want to get permission up front). I'll probably post my summary to the sci.nanotech and the sci.physics newsgroups. (Perhaps even to sci.optics if many of the devices are optical).
Thank you for your time.
-- David Cary d.cary@ieee.org.
Newsgroups: sci.physics From: bhv at areaplg2.corp.mot.com (Bronis Vidugiris) Subject: Re: Quantum Computing? Organization: What - me organized? Date: Wed, 18 Oct 1995 15:21:55 GMT ... David H. Lee <davelee at davelee> wrote: ) )I read a post somewhere recently that stated that 'Quantum Computers' )will be able to solve currently unsolvable problems. ) )However, this person failed to define 'Quantum Computers'. ) )As far as i know, current microprocessors must take quantum-effects into )account because of the sub-micron structure of the circuits. ) )What exactly is a 'Quantum Computer' that makes it so special? )Are these computers that untilize 0 dimensional quantum-boxes and 1 )dimensional wires to speed up computing by several orders of magnitude? )Or do they simply exploit quantum mechanics with conventional )semi-conductor technology? If you have a www browser, you might want to try: http://vesta.physics.ucla.edu:7777 or the newer http://vesta.physics.ucla.edu/~smolin/ (I've often had better luck with the link that's supposed to be out of date, go figure). I'd particularly recommend the paper by Shor, you can find this in print now as well (proc 35th annaul symposium on the foundations of computer science). Quantum computing is based on the manipulation of quantum state vectors by unitary transformations - actually "small" unitary transformations. Quantum state vectors grow in length exponentially as one adds "qbits", 2-state quantum systems. It takes 2^n complex numbers to represent 1 qbit. This suggests one reason why there is no known way to simulate quantum systems in less than exponential time on a classical computer - the mere state description grows exponentially! "Small" unitary transformations are the tensor product of a "small" matrix (2x2, 4x4 or somesuch) with an identity matrix of some appropriate size to get the required dimensionality to match the dimensionality of the state vector.
From: caj at sherlock.math.niu.edu (Xcott Craver) Newsgroups: alt.security.pgp,sci.crypt,sci.physics Subject: Re: Quantum computers--what would it take? Date: 11 Oct 1995 04:31:41 GMT Organization: Northern Illinois University, Math ... Jon Leonard <jleonard at solcom.eng.sun.com> wrote: ... >David Sternlight <david at sternlight.com> wrote: ... >>Ed Falk <falk at peregrine.eng.sun.com> ... >>>If somebody invents a real-live quantum computer, your life will >>>be changed so radically, and at such a fundamental level, that the >>>fact that PGP is broken will be the least of your worries. >> >>Chuckle of the day. Indeed. "Wait, 'quantum' - that's, like, a physics BUZZWORD, right? Like on Star-Trek, RIGHT? So we could like, invent transporters and fly faster than light and build positronic brains and-" BLAMBLAMBLAM ... thud. >Some real-world examples of things that would change: >Compilers could magically produce optimal code quickly. >Airplanes would be built using the perfect aerodynamic design, instead >of the best a team of engineers could come up with. Erm, maybe, but I certainly don't consider that such a radical change, "and at such a fundamental level," that using it to break PGP is negligible in contrast. As far as radical changes go, breaking public-key encryption systems is a big, big deal. It's a matter of both constitutional rights and National Security. I invite you to think of an application of computer science that would make cracking RSA 'the least of our problems'. Furthermore, the difference between Quantum and non-Quantum computing is *NOT* <===!! the difference between NP and P!!! Quantum algorithms for performing NP-complete problems in reasonable time have not yet been found. Factoring, remember, just may be P, even if P != NP. Also, don't confuse NP-complete with EXPTIME. Even if you could do NP stuff on your brand-new Quantum machine, there would still be a good body of important problems you still couldn't solve. In other words, no miracles. But back to the original subject: namely, what technological constraints keep quantum computers (quputers?) from being built, and how long do you think before we overcome them, if they can indeed be overcome? >Jon Leonard -Caj wants a quantum BeBox! YOW! -- 88 ,ad8888ba, "" <=-- That's right - the '@' character is actually MY NAME, d8P' `"':88 <=-- reduced 10.7 times for transmission purposes. Forget d8: ;88 <=-- some stupid name like "octothorpe" or "virgule": the 88: ,adPYa888 <=-- '@' is hereby officially called the CAJ!!1! All that 88: 88 `88 <=-- time you spent programming in FORTH you were actually Y8: 88, ,88 <=-- aiding, unwittingly, the propagation of my *almighty* Y8a. `Ybaa8P88 <=-- *wisdom*! KIBO may have a newsgroup and a stupid lil o `"Y8888YP"d8P <=-- dog, but I RESIDE IN ASCII!!!!! I AM IMMORTAL!!!11!! Yb, _,8P' `YY88888888P"' Caj@math.niu.edu -- My opinions do not represent.
[FIXME: move other photonics references here]
I suspect there is a time period where most computation is photonic ... a narrow time wedged between CMOS electronics and early nanotech.
See also machine_vision.html
[FIXME: ...]
http://www.thehub.com.au/~mitch/extro/nano.html#dna (DNA computing)
"rod logic": Drexler described mechanical computing systems in _Nanotechnology_.
http://www.thehub.com.au/~mitch/computing.html (future computing)
``Brainy Bacteria'' article by Dolly Setton, Best of The Web, 09.10.01 http://www.forbes.com/best/2001/0910/016.html
[FIXME: bignums#money]`` One appeal of bacterial computing is that bacteria are very cheap to manufacture. Scientists can grow trillions of bacteria in a lab for a few dollars in material costs. ... It's an appealing prospect, given how expensive computing is today. It costs about $2 billion to build a new semiconductor fabrication plant, where people in space suits work with robots to build the chips that go into personal computers. ''
"chaos-based computing" http://www.gtri.gatech.edu/res-news/DYNAMIC.html
What is a Computer? ftp://ftp.ans.net/pub/misc/jta/biocompflow.txt by John-Thones Amenyo <jta at ans.net> Copyright 1995, J. Amenyo Lots of various computing possibilities (optical computer, fluidic computer, biocomputers, quantum computers, etc.)
"Plasma waves" http://www.rpi.edu/dept/NewsComm/Review/sept97/sept_12/waves.html
An introduction to Quantum Computing, why it's nifty, quantum error correcting codes, etc. http://www.lucifer.com/~sean/BT/14.html#ClarkQuantum
It's possible that one application of quantum computing will be towards Computed Holography.
The Quantum Optics and Atom Optics links page http://www.physics.mq.edu.au/~drice/quoptics.html
Date: Mon, 2 Dec 1996 00:18:02 -0500 From: Michael Frank <mpf at ai.mit.edu> To: David Cary Subject: Re: quantum computing links X-URL: http://www.ai.mit.edu/~mpf David Cary writes: > > Would you mind me adding a link to your page on my quantum computing page ? Sure, no problem. > > You have some nice content on your web pages. Perhaps you would be > interested in the quantum computing page on my web page. > Thanks, your page looks nice. Interesting to see people discussing the relevance of quantum computers to public-key cryptography. In my view, this is still very much up in the air, even if good quantum computers could be built, because computer scientists know of plenty of other, equally secure public-key cryptography schemes that have nothing whatsoever to do with factoring. The only reason RSA is so popular is that it was first. But if factoring fell, one of the other cryptosystems could just rise up and take its place. Of course, it's possible quantum computers would be able to solve those other public-key cryptosystems as well, but that would have to be demonstrated. But anyway, there would still be a big short-term impact while people switch their software over from RSA to the other non-factoring-related public-key cryptosystems. -Mike > p.s.: > On your page http://www.ai.mit.edu/~mpf/qclinks.html, you have > -- > ADD_DATE="818813634" LAST_VISIT="818821896">Q-gol Quantum Programming > Language > -- > which has moved. > > David Cary <*> O- > Future Technology. > >
"Everyone has some intuition about baseball and billiard balls [which is invaluable for engineering mechanical devices. ...] But very few people have intuition about these quantum-mechanical interference and information effects. And the only way to make anything useful with them is if we really do have that good intuition." -- Los Alamos National Laboratory physicist Paul Kwait, quoted in _Science_ 1996 Oct 25 p. 505 article "To Send Data, Physicists Resort to Quantum Voodoo".
[FIXME:]
Date: Tue, 10 Dec 1996 From: John Douglas Gleason <gleasojb at bc.edu> To: transhuman at logrus.org Subject: Re: >H The Singularity and the Impossible ... Check out the October 11, 1996 issue of _Science_. According to Paul Parsons (University of Sussex) faster than light travel is theoretically possible with general relativity. What makes it even more interesting is that it also allows for travel back in time. Though it may be theoretically possible, that doesn't mean we will ever be able to do it. There's the old question, if it's possible to travel back in time then where are all of the tourists from the future? A _Scientific American_ article a couple years back presented the parallel universe argument that they just haven't shown up in our time-line. Interesting stuff to discuss with physicists over a few beers! As long as I'm promoting _Science_ magazine <g>, the Oct. 25, 1996 issue has a neat story on some experiments with quantum communication and teleporting. It also mentions possible applications to quantum computing. The full text is available online at http://www.sciencemag.org/science/scripts/display/full/274/5287/504.html John
Date: Tue, 28 Jan 1997 10:20:19 -0800 (PST) From: John K Clark <johnkc at well.com> To: transhuman at umich.edu Subject: >H Quantum Computers Reply-To: transhuman at logrus.org Transhuman Mailing List -----BEGIN PGP SIGNED MESSAGE----- Very recently there have been two new developments in the world of Quantum Computers. In the January 17 1997 issue of Science is a long article called "Bulk Spin-Resonance Quantum Computation" by Neil A Gershenfield and Issac L Chuang, there is also an editorial about the article in the same issue. It talks about hardware methods of isolating a Quantum Computer from the environment so you don't get Quantum decoherience and suggests using Nuclear Magnetic Resonance (NMR). The nucleus of atoms are like spinning bar magnets, and they can be in 2 states, spin up or spin down. If you send in a radio pulse of the exact frequency you can change the spin state of the atom from one to another. This frequency can change if an atom is coupled to another atom, so if you pick the correct frequency you could flip the spin of a carbon atom if and only if the Hydrogen atom it is coupled with is spinning up. This could make a controlled logic NOT gate. This would also be a Quantum logic gate, because a radio pulse can also put an atom in a superposition of states, both spin up and spin down at the same time, and anything coupled to that state would also be in a superposition of states. There are 4 advantages to using NMR to make a Quantum computer rather than other methods: 1) The quantum state of interest is in the nucleus and is protected from contamination from the outside world by the electron cloud so the quantum state lasts a long time. In most other quantum systems decoherence happens in pico or femtoseconds but with NMR it can last for thousands of seconds. 2) You don't have to worry about every single atom, if only 50.000001% point up, that's good enough. I quote Tim Havel, who independently discovered most of the thing in the article at the about the same time, "It turns out having a well-defined statistical excess of a single quantum state is enough to do quantum computing. You don't need to have every molecule doing exactly the same thing". 3) By using a large number of atoms to store a qbit it further protects it against unwanted external interactions. 4) You don't need expensive exotic gadgets, cheap off the shelf equipment will do for NMR. The authors actually hope to have a small (10 qbit) quantum computer in operation in a few years, possibly as soon as next summer. With such a machine you could factor (drum roll please) the number 15. Hey don't laugh, it would be a proof of concept and could serve as a test bed to develop better quantum algorithms, and remember, each time you add a qbit to your computer you double its power. Speaking of better quantum algorithms, we'll need some to go much higher than 30 qbits using NMR, and to quote Gershenfield "Somewhere between 50 and 100 qbits is where calculations start to get interesting". Still, people are optimistic, to quote Gershenfield again "If researchers can figure out how to extend the technology to larger qbit numbers the future is unlimited. In a separate development, on January 27 researchers at Xerox announced that they have made a small molecule of manganese, oxygen, carbon and hydrogen that acts like a powerful magnet. They said it might be possible to use this material to make a disk drive that could store information at the molecular level and hold millions of times as much data as we can today, but that's not what they emphasized. I think it's interesting that at the announcement they spent more time talking about how it could be used in a Quantum computer rather than a conventional one. The material has a property called Quantum Magnetic Hysteresis, it can be in a superposition of many different magnetic states at the same time, and that's just what a Quantum Computer is all about. For me It's still hard for me for it to sink in that an otherworldly object like a Quantum computer could actually exist in our worldly universe, but it's beginning to look like one can really be built. If so the world will never be the same. John K Clark johnkc at well.com
Date: Fri, 31 Jan 1997 09:38:36 +0100 (MET) From: Anders Sandberg <nv91-asa at nada.kth.se> To: transhuman at logrus.org Subject: Re: >H Transhumanistic Companies/organizations/groups/individuals/etc. X-MIME-Autoconverted: from 8bit to quoted-printable by siddhartha.us.itd.umich.edu id DAA12815 Transhuman Mailing List On Fri, 31 Jan 1997, Jos Sez wrote: > > Uploading: there is plenty of research going on about simulating realistic > > biological neurons. > How do you do that? I want DETAILS. Some more details (but still incomplete): Most work have been done in simulating the electrodynamics of the membrane and synapses by using models for the ion channels and their interactions. Usually one divids a simulated neuron into cylindrical or spherical compartments (asumed to be isopotential) with membrane properties set by the concentrations of different ion channels. Then the membrane polarization and state is calculated by solving differential equations (based on the Hodgson-Huxley equations) where each compartment influences the neighboring compartments (essentially everything can be seen as a electronic network). Several such model neurons are connected by synapses, and one studies the output or behavior of the network. Of course, the big problem is (a) making a good membrane model, and (b) doing the number crunching (we are happy here at the Institute that we got several paralell computers). See for example the pages of the SANS group here at Nada (http://www.nada.kth.se/sans I think) for more discussion. There are links from there into the wide world of neural networks (although most people deal with artificial NNs). [A small aside: the quantum computer discussions gave me a fun idea, quantum neural networks. It seems that you can train a quantum backprop net in just one iteration if you can feed it a superposition of all training patterns and compare the output with a superposition of all desired output... I think.] > Since nobody said anything about Si based proteins I assume they don't > exist as I suspected (Never liked the idea anyway). I don't think they do. Silicon tends to form weaker bonds than carbon, so it is possible that you can't make the peptide backbone stable. ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! http://www.nada.kth.se/~nv91-asa/main.html GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
quantum computer http://www.nytimes.com/library/national/042898quantum.html
"John Preskill has a beautiful set of [quantum computing] lecture notes available" http://www.theory.caltech.edu/people/preskill/ph229/
Quantum Computers http://qso.lanl.gov/~gottesma/QComputers.html
Date: Tue, 24 Feb 1998 21:23:12 +1000 (EST) From: Mitchell Porter To: transhuman at maniac.deathstar.org Subject: >H re: Topics of current personal interest Reply-To: transhuman at logrus.org Transhuman Mailing List (I get the digest, so I'm seeing this weeks afterwards.) David Cary said > >2. Developing useful applications of NT. > ... > >so > >that we're ready to use the technology the moment it > >arrives. > > Good point. The most practical plan I've seen so far is to make better golf > clubs (!) as part of the early bootstrap process. I don't remember that one! Where'd you see that? > >5. Conceptualizing and planning for "infinite futures". > >For example: > > > >One can imagine a very fast computer connected to > >sensors and effectors throughout a baby universe. > >The computer maintains a model of the state of the rest > >of the universe, constantly updated via the sensors. > >There is always a particular finite range of actions > >open to it through the effectors. What it does is to > >model the effects of each possible action, some degree > >into the future, and selects an action on the basis of > >some criterion ('value system'?) applied to the various > >projected outcomes. > > How is this different from what sentients do already ? The difference would be one of degree. I'm talking about a situation in which the baby universe is saturated with sensors, in which its model is constantly updated in realtime, and in which simulations of its future can run faster than the real thing. So it's a situation of effective automated omnipotence (if you throw in some actuators as well, that is). I'm curious about the logistics of such situations (e.g. under varying cosmic scenarios: finite versus infinite universe; baby universes are possible vs black holes are just massive elementary particles and don't "lead" anywhere), whether such a future is likely or necessary, how unreliability of the supervising system or unpredictable future changes in the intentions of its programmers might affect the picture, whether we have any values which lead us to find such a future inherently undesirable, and so on. > >10. Quantum compiler. We need a way to turn programs > >into something that can be run on a quantum computer. > > Should we work on this before or after building quantum computers capable > of running these programs ? Thinking about it now should be useful, but first you have to catch up with the state of quantum computational theory, and then you have to make it make sense. I'm told that just about every reasonable position regarding the relative powers of classical and quantum computers has its proponents, for example. See http://www.theory.caltech.edu/people/preskill/ph229/manifesto5.ps for one "manifesto". Q-gol, at http://rosebay.matra.com.au/~gregb/q-gol//index.html, is supposed to be a quantum-computing programming language, but I can't connect to that page any more. A web search on "quantum compiler" turns up http://www.darpa.mil/ito/Summaries97/F361_0.html, which lists a string of goals including development of a quantum compiler and of a desktop quantum computer. That page is about NMR quantum computing, which is really classically-parallel bulk quantum computation, in which an Avogadro's number of molecules are used as individual quantum computers, and NMR is used to apply the quantum logic gates and to read out the final state. I'm intrigued by the notion of solid-state NMR quantum computing. NMRQC has the advantage that the output of a quantum computer is generally probabilistic (e.g. Shor's algorithm) and here you're automatically averaging over the output of a zillion quantum computers running the same program. But all those molecules are tumbling around and drifting about and undergoing thermal interaction - surely it would be advantageous to have your "single-molecule quantum computer" in a vacuum and fixed in space. So I imagine using advanced nanotech to construct some rigid, perfectly periodic structure containing ultra-high-vacuum spaces in which identical copies of the SMQC exist... > ... > >12. What it's like to be 'auto-omniscient' (knowing > >everything about oneself) and 'autopotent' (having > >the capacity to change any aspect of oneself - term > >due to Nick Bostrom). > > Deep questions. > > I believe A. Sandberg claimed that the free will is caused by *lack* of > knowledge about oneself. 'Free will' is one of those terms that *has* to be defined before you can hope to converse with any mutual understanding. I've heard people say that one is free insofar as one's behavior is brought about by one's conscious knowledge and desires (and not external or unconscious influences), in which case freedom is increased by self-knowledge. -mitch http://www.thehub.com.au/~mitch[FIXME: #free-will]
Quantum Computing: QUIC Theory at Caltech http://www.theory.caltech.edu/~alandahl/quic.html [offline ?]
"Analog computer trumps Turing model" By Sunny Bains http://www.eetimes.com/story/OEG19981103S0017 "Recent developments ... suggest that analog computations are more powerful than digital ones."
RSFQ Laboratory http://pavel.physics.sunysb.edu/RSFQ/ "RSFQ stands for Rapid Single Flux Quantum logic/memory family. This is an emerging superconductor technology that promises digital data processing at clock frequencies of up to 600GHz with negligible power consumption of ~0.4uW per Josephson junction." Apparently researchers here have already attempted to build a complete general-purpose RSFQ microprocessor. Special software here.
Introduction to Nonlinear and Chaotic Phenomena http://trixie.eecs.berkeley.edu/~chaiwah/introduction.html (Chua's Circuit)
http://dir.yahoo.com/Science/Physics/Quantum_Computing/
"The QUIC web page has been taken off-line. For current quantum computing and quantum information research at Caltech, please start at one of the following sites:
"
From: Mitchell Porter <mitch at thehub.com.au> Subject: tech: [Velcro] DARPA - UltraScale Computing (fwd) To: transhumantech at excelsior.org Date: Thu, 11 Jun 1998 17:47:46 +1000 (EST) Reply-To: transhumantech-l at excelsior.org This is a sort of glossy overview of various new forms of computation... ----- Forwarded message from Andrew Wood ----- Have a look at what the US Defence computing researchers are writing pretty reports about... http://www.darpa.mil/ito/related/maynard.pdf includes such great terms as: swarm computing, dna computing, celular engineering and orchestrated organisms. And of course, no defense report would be complete without a quote from the master himself: "A clever fighter is one who not only wins, but excels in winning with ease" Sun Tsu Andrew. --- The second law of free assortment states that in a cross involving one pair of alternative characteristics, the characteristics will segregate in the second filial generation in the realtive proportions of 9-3-3-1. ----- End of forwarded message from Andrew Wood ----- Send unsubscribe requests to: majordomo@excelsior.org Archive located at: http://www.excelsior.org/transhuman_tech_list/
Princeton - August 23, 2000 - Could bizarre giant particles nearly as big as galaxies, but lighter than electrons, dig cosmology out of a hole? Physicists suggest the Universe is jam-packed with these strange particles, which might explain why small galaxies are far rarer in the Universe than theory predicts.
"I know it sounds crazy," says team member Andrei Gruzinov of the Institute for Advanced Study in Princeton, New Jersey. "But so far as we can tell, no observations rule it out."
>Date: Thu, 13 Aug 1998 09:05:48 -0700 (PDT) >From: Michael Nielsen <mnielsen at theory.caltech.edu> >To: transhuman list <transhuman at logrus.org> >Subject: Re: >H quantum computers : a naive question >Reply-To: transhuman at logrus.org > >Transhuman Mailing List ... >On Thu, 13 Aug 1998, Remi Sussan wrote: > >> I have read somewhere that quantum computers were impossible , in fact >> the decoherence effect (same word in english ?) would kill quantum >> incertainty, because what happens at the micro-level disappears at the >> macro one. > >This is a possibility which has been extensively investigated. Since then, >the effects of errors, and how to minimize that effect, has been >understood much better, and it seems that there is no fundamental reason >why quantum computers can't be made that are resistant to the effects of >noise. > >The result which shows that quantum computers can overcome the noise >problem has become known as the "threshold theorem". The result is due to >the joint work of many people over a period of about 2 years. It states >that provided logic gates can be implemented with an error rate below some >"threhsold" value (current best estimates are about 10^{-4}), then it is >possible to make a quantum computer arbitrarily accurate, for arbitrarily >long computations. > >A fairly readable account of the threshold theorem may be found in >Science, earlier this year. The title of the article is "Resilient quantum >computation", and the authors are Knill, Laflamme and Zurek. Sorry that I >don't have a more detailed reference -- I'm moving at the moment, and my >computing power is all over the place. A longer article, but perhaps >even more readable, is by Preskill, and appeared in the February (?) 1998 >issue of the Proceedings of the Royal Society of London A. (That was a >special issue on quantum computation, and contains several >useful survey-level articles.) > >> As a quantum computer would -obviously- have some >> interactions with its environment, the quantum "bits" would finally >> adopt only the classical two states yes-or-no common in current >> computers.(well this is perhaps a rather crude explanation). > >That's actaully a pretty nice way of saying it. Other error processes >beside the one you mention are important as well, but that "classicizing" >error is one of the most important. > >> I see people on this list convinced that quantum computer are not only >> possible, but probable and they probably know the field much more than I >> do. > >A tremendous amount of work remains to be done before such an assessment >is justified, in my opinion. > >Michael Nielsen > >http://www.theory.caltech.edu/~mnielsen/
Bernhard Omer... has written a high-level, architecture-independent programming language for quantum computers. ... called QCL, short for "Quantum Computation Language," ... http://tph.tuwien.ac.at/~oemer/qcl.html
Date: Mon, 16 Feb 1998 11:52:25 -0800 From: Ron Blue <rcb1 at lex.lccc.edu> MIME-Version: 1.0 To: d.cary at ieee.org Subject: quantum robot 3.24 trillion qubit robot at http://www.neutronicstechcorp.com/private opporational August 1996[FIXME: offline ? I am skeptical this ever really existed ...]
-- http://lenslet.com/ "World's First Commercial Optical Processor by LENSLET"Lenslet, a leader in Optical Digital Signal Processing (ODSP), is developing the first commercially available Optical DSP capable of up to 8,000 Giga MAC operations per second.
Lenslet's ODSP enables the emergence of computation-intensive applications including Multi User Detection, Smart Antenna, Software Defined Radio; real time H.264 video compression for commercial, industrial, military and homeland security markets.
Send comments, suggestions, bug reports to