The D-Wave company announced in june a processor of more than 1000 quBits. They said that "At 1000 qubits, the new processor considers 21000 possibilities simultaneously,.... In fact, the new search space contains far more possibilities than there are particles in the observable universe." It seems that Google and other important companies are interested in the project.
What do you think? Quantum computing is ready to commercialise or just a commercial trick for getting investment?
http://www.dwavesys.com/press-releases/d-wave-systems-breaks-1000-qubit-quantum-computing-barrier
We should give credit to D-Wave, their technology is amazing, but it was never meant to be a universal quantum computer. The chip does adiabatic quantum optimisation with a restricted qubit topology. The presence of thermal noise and external magnetic fields means that the system is not fully coherent, furthermore, the annealing time does not obey theoretical bounds. Last year more than two dozen papers appeared an arXiv that studied what the hardware could do and how fast: now we are confident that quantum tunnelling plays a role in finding low-lying excited states, and some forms of entanglement may also be present. In short, the D-Wave chip is able to find a good local optimum for certain kinds of nonconvex objective functions, which makes it interesting for machine learning applications and certain engineering problems.
Unfortunately, a fully coherent universal quantum computer remains a tough challenge. On the other hand, many other companies are also investing in adiabatic technologies, which, by the way, is equivalent to a universal gate-based quantum computer with a polynomial overhead if it has complete qubit connectivity. Encryption cracking is another pathway, but that is mainly interesting to government agencies and not to private companies.
We should give credit to D-Wave, their technology is amazing, but it was never meant to be a universal quantum computer. The chip does adiabatic quantum optimisation with a restricted qubit topology. The presence of thermal noise and external magnetic fields means that the system is not fully coherent, furthermore, the annealing time does not obey theoretical bounds. Last year more than two dozen papers appeared an arXiv that studied what the hardware could do and how fast: now we are confident that quantum tunnelling plays a role in finding low-lying excited states, and some forms of entanglement may also be present. In short, the D-Wave chip is able to find a good local optimum for certain kinds of nonconvex objective functions, which makes it interesting for machine learning applications and certain engineering problems.
Unfortunately, a fully coherent universal quantum computer remains a tough challenge. On the other hand, many other companies are also investing in adiabatic technologies, which, by the way, is equivalent to a universal gate-based quantum computer with a polynomial overhead if it has complete qubit connectivity. Encryption cracking is another pathway, but that is mainly interesting to government agencies and not to private companies.
The number of qubits doesn't say anything to the lay man. What we want to know is what this computer is capable of. The main problem is decoherence, which makes the number of possibilities rapidly fall to the one of the classical computer. As far as I know, the best performance either is to have found out that 15 = 3 x 5.
- Badges
- Science topic
- Similar topics
- Engineering
- Computer Engineering