Google
Support independent publishing: buy this book on Lulu.

Saturday, July 27, 2019

QUANTUM COMPUTER







A BREAKTHROUGH IS COMING

"Quantum computers can keep up with the (true), complexity of nature." “Accurate modeling, has applications for engineering, medicine, energy production and much more”: Richard Feynman


Although a quantum computer (QC), prototype already works and is marketed  (D-Wave S.), installed in modules of 5 x 5m and heights of 7m., insulated with fiberglass from the environment, with 50 qbits inserted in silica receivers, protected by a near zero absolute cooling, with softwares adapted to electrons, with environmental insulation, to avoid interference and errors. What is missing ... What is left over? There is an excess of optimism, such as that of Geordie Rose, co-founder of D-Wave System, who has made some predictions: One) 2020, using a QC, the NASA will identify 40 light years from Earth, a planet with atmosphere and oceans of liquid water, similar to the Earth. Two) 2023, Experiments with QCs, will demonstrate the existence of parallel universes. Three) 2028: QCs will play a critical role in creating robots, which will match everything a human does. On the other hand, Google, Intel and IBM, also have prototypes of QCs, not commercialized of 45 qbits, although they hope to escalate  the number of their qbits to 1000 or more, in order to be able to predict climatic changes, perform data encryption (100%, safe), develop new materials, etc. What is missing? 1) Avoid decoherence (qbits collapse), induced by noise, temperature changes, vibrations. Given this fragility, the qbits, require temperatures close to absolute zero (-273 oC Celsius), to remain stable. 2) New software and interconnect technologies are needed to take advantage of the gigantic processing power of QCs. As it is known while in classical computers (CC), a bit (transistor), supposes 2 separate states (0 and 1), in a  QC, a qbit is 0 and 1 at the same time (superposition), a phenomenon that together with entanglement (combinations of qbits  superpositions), allow a QC, to calculate in hours the  complex modeling  of a molecule, which possibly would take years to a CC. 300 combinations of exponential superposition  of qbits would produce more overlapping states than existing atoms in the visible universe. 3) There are 6 to 7 different types of qubits, although only 3 or 4 are used. To obtain a superposition, 2 qubits are needed and for an entanglement more than 3. Some qbits need superconductors (where electrons flow without resistance), others: oscillating charges of ions kept fixed by lasers, also silica qbits with a single electron, controlled by microwaves. 4) Increase the number of qbits to 1 million. Continue working with QC simulators. Increase the number of math softwares. Improve the accuracy of qbits controller microwaves. 5) Given the enormous amounts of money to be invested, it is important to have a coordinated work strategy between universities, governments and industry. 6) Promote the growth of artificial intelligence (AI), through the use of QCs. 7) Establish the supremacy of   QCs over the CCs, solving problems that the latter cannot do. Establish new ways (faster, more efficient), to perform certain calculations, in order to facilitate modeling of the brain, of complex molecules, of factoring large numbers solving encryption problems, taking as an example the Shor's algorithm (factorization of large numbers), of great practical contributions. According to Scott Aronson (Texas University), QCs must solve certain problems, better than CCs. It is important that QCs can solve the random circuit sampling problem (take random efficient quantum circuits and generate samples for external distribution), build quantum circuits of at least a certain minimum size, with which CCs, do not count. 8) Reduce the error, especially now that the number of gates and qbits must be increased. The most crucial mistake is that which occurs every time a gate is used. Try that error for gates of qbits, be around: 0.1%. 9) How much is missing to have a universal quantum computer? Google that favor the supremacy of QCs, hopes to present one of them, at the end of 2019. Without ruling on the date: IBM, IonQ, Rigetti and Harvard University, say it will be very soon. Harvard University that   uses rubidium atoms and Microsof atoms that uses topological qbits, need more time. Since silicon qbits are fixed and need to be individually calibrated,they  will have a difficult task at  time of increasing the number of their qbits. Trapped ions give more time to make sure of their collapse, due to environmental noise. On the other hand, its gates are very slow allowing ions to move, when they are not needed. Some think that   qbits should adopt 2 bit qualities of CCs: be easily scalable and impeccable. 10) Anything else? According to Adam Bouland (California, University, Berkeley), after QCs solve the problem of random circuit sampling, there will be others, especially everyday tasks: financial services, AI, chemistry, creation of  QCs capable of self-eliminating their errors in real time, so that at  the end, they will  offer calculations without errors, which will require massive amounts of error-correcting qbits, connected to each logic gate.



Labels: , , , , ,

Wednesday, July 24, 2019

POTENTIAL TELEPATHY?







IT'S NOT TELEPATHY, BUT IT IS NEAR TO BE

Telepathy: "Supposed communication of thoughts or ideas by means other than the known senses."

If for some reason we were conscious but unable to speak, how would we communicate with the rest of the people? Recent research conducted at the University of San Francisco (UCSF), by Josh Chartier and Edward Chang, allows the generation of synthesized spoken language, from about 100 brain signals responsible for precisely coordinating the movements of lips, jaw, tongue and larynx, inducing our breathing to form words and sentences. Only the medium used (electricity) separates the definition of telepathy from what has been achieved. 5 volunteers who were being treated for epilepsy read phrases aloud, while a grid of 256 electrodes placed on the surface of their cerebral sensory motor cortices recorded and measured the resulting signals. Computational models based on these data allowed researchers to decode the way in which the patterns of activity of speech brain centers contributed to generate particular movements in their vocal tracts. Later, the simulated movements of the vocal tracts were transformed into sounds, generating intelligible synthesized speech. The sound patterns of the individual words synthesized from brain activity were very similar to those originally spoken. The model can also decode the acoustic differences between participating voices and can also do so even when the participants mouthed sentences. Now, enormous possibilities are open.



Labels: , , ,