is the practical use of quantum computers within reach?
2024-10-05
한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina
reference news network reported on october 5 that according to a report on the british "new scientist" weekly website on september 30, some recent experiments indicate that researchers may be able to develop a useful quantum computer as early as 2029.
according to the report, the researchers said these latest experiments represent a key milestone and mark that we are entering a new era. "suddenly, really useful devices seem within reach like never before," says jamie vicari of the university of cambridge in the uk.
over the past decade, quantum computing companies have mostly focused on building larger and larger devices, gradually increasing the number of qubits in the system. qubits are the basic units of quantum information, realized through physical systems such as electron spin or photon polarization. but qubits are prone to errors and cannot reliably run algorithms used in the real world.
now, the companies appear to be turning their attention to building error-free qubits, known as logic qubits. these physical qubits can reduce errors to low enough levels to run algorithms consistent with the real world.
"this marks a reset and a new benchmark for everything around quantum computing," vicari said. "it's exciting because this is the moment when quantum computers start to become useful."
in august, google researchers published a paper showing that when building logical qubits by adding more physical qubits to a computer, errors don't snowball out of control but cross a threshold. after the leap, in principle, errors will decrease as the system scales. the idea is to spread the information across a group of qubits, so if an error occurs in one qubit, it doesn't affect the overall calculation.
dan brown of university college london said: "the ideal situation would be to further reduce the error rate and improve the quality of the qubits while scaling up. doing these things at the same time is very difficult, but what makes me optimistic is that google's experiment proof that they are trying.”
however, roberto bondesan of imperial college london said that google's work does not involve performing calculations through qubits - instead, the researchers showed that these qubits can act as memory.
another study published in september by researchers at microsoft and quantum computing startup quantitative quantum computing combined error-correcting qubits and computing. the team set up different combinations of qubits, creating four logical qubits, and then performed basic logic operations on the system, such as turning the qubit's value from positive to negative.
"they have fewer rounds of error correction, so their quantum memory has a shorter stabilization time, but it can also be used to do some calculations," bondesan said.
the quantum computer in microsoft's research uses a different hardware design than google's, using a series of magnetically trapped charged particles rather than superconducting wires.
this allows it to use an error-correction technique that arranges qubits in a complex geometric structure (a four-dimensional hypercube). "in principle, they can encode more logical qubits with fewer physical qubits. in that sense, it's more efficient," bondesan said.
other researchers have shown how error correction can be done in more alternative quantum computers. benjamin brock of yale university in the united states and his colleagues tested an error-correction technique called bose coding, which distributes errors across the vibrations of a quantum computer. this system uses not qubits, but "high-dimensional qubits," which can take on more values than 1 and 0 and are theoretically more powerful. in addition, researchers from amazon's quantum computing team demonstrated another type of bose encoding called "categorical qubits," which, similar to google's results, show that as the system scales up, its errors decrease.
"the google and microsoft approaches follow more mainstream qubit-based quantum computing, while the bose encoding introduced in the yale and amazon experiments is more novel and exploratory, but also very exciting," brown said.
brown said the error correction worked across a number of different designs and experiments, suggesting that previous theoretical results could apply to real-world systems. he said: "there are a lot of good theoretical ideas about fault-tolerant quantum computing and error correction, but none of them have been proven, or they have been proven in a very limited way or restricted to special cases. things have really changed a lot in the past few years. changes. we hit a number of key milestones, demonstrating different aspects of fault-tolerant quantum computing.”
however, the pace of progress may be hindered and a fully fault-tolerant system may remain out of reach, which could dash the hopes raised by google and ibm's optimistic timelines. "i would be surprised if it goes very well," brown said. "i think there will be barriers. each of these platforms will scale as far as they can until they hit a barrier. it's hard to predict where the barriers will arise." ( compiled/xiong wenyuan)
