Major improvement: So far, quantum computers have been prevented from being used on a large scale due to errors. Experiments with Google’s quantum computer “Sycamore” have now demonstrated in practice that such correction protocols work for the first time. According to the team in the special issue of “Nature”, they reduced the error rate of the system by 100 factors. But more important than that: error correction when measuring has a high-speed effect, which only makes large quantum systems possible.
Quantum Computer Are considered to be the computers of the future because they are much faster than ordinary computers because of quantum physics phenomena such as problem and superposition. Google Calculator “Psychmore“And Chinese Quantum Simulator”Jiujang“This quantum must have already proven superior. However, technology is not yet fully developed: quantum computers still make many mistakes.
Why there are so many mistakes in quantum computing
Problem: The complexity and superposition of quantum bits are so easily affected that even the smallest confusing impacts can end their synchronization and lead to errors. Quantum computers can only be used reliably and on a large scale if there are corrective methods to effectively detect and eliminate these errors. Some such algorithms have already been developed, but their practical testing has not yet been carried out on more complex quantum computers.
A team of Google AIs is now conducting these experiments using their quantum computer “Sycamore”. The system is based on 54 superconducting quits, which are arranged in two dimensional phase and connected to each other. The error correction principle is based on combining several physical quidets into one logical quid that ultimately does the calculations. Pros: This means that the calculation can continue despite individual failures of the body units, while some quid can be used for error monitoring.
Two error logs in the test
Unlike a normal computer, quantum bits cannot be copied because a quantum system cannot detect errors by copying and comparing data. So we need to find a way to find the errors without destroying the quantum state. Such error monitoring is possible because individual quidets detect the quantum state of their neighbors and thus act as “measuring sensors”.
The Google team has now tested two such methods with their “Sycamore” system. First repeat code. “Again with indexing, individual quidets switch back and forth between measurement and data processing in one dimensional chain,” they explain. “Each measurement checks the status of its two neighbors.” However, only one of the two types of errors per step is detected – a bit or a build error.
In the second method, the surface code, measurement and data quid are alternately arranged in a two-dimensional phase. Each measurement can detect both quid bit and phase errors. So this method is considered to be very suitable for large quantum computers, but it is very difficult to implement. In their experiments, the researchers again executed the code in chains up to 21 quid and the surface code in small quantities of seven quits.
Correction works exponentially
Conclusion: “Again, if the number of audits is increased from five to 21, the logical errors are reduced by more than a hundred times,” the scientists said. This suggests that error correction is progressing rapidly with increasing quid. Such high-speed debugging is important, otherwise the effort on large quantum systems would be unacceptably high – the benefit of the quantum system would then soon become obsolete.
“Theoretically the error correction rate was expected to increase exponentially with the number of seconds used. However, this theoretical hypothesis can be seen to be fulfilled.
“An Important Milestone”
Also positive: the practical implementation of the interface code was also successful, and the system behaved as predicted in the samples. The correction methods were consistent for more than 50 cycles, the research team reported. “These test demonstrations now lay the groundwork for the creation of a measurable, fault-tolerant quantum system based on superconducting quid,” the scientists write.
Sven Ramelo and Helen Krsanovsky of Humboldt University in Berlin also see this as an important turning point: “One of the undisputed world leaders in the race to design the first global quantum computer has taken one of the most important milestones along the way.” “Although it is still a very small device compared to the view of a real, effective quantum system, it shows for sure that effective and above all efficient debugging is practically possible on a quantum processor.”
There are challenges
However, before quantum computers can be calculated on a large scale without error, some optimization is still necessary, which the Google AI team agrees. “There are still many challenges on the way to measurable error correction,” they write. Above all, to use the surface code, quid synchronization and control must be further improved.
Quell: Center for Nature and Science Media