**Quantum computer experiments at UMD show that combining quantum computer fragments does not have to mean combining their error rates.**

Pobody’s Weaknesses — The indifference and even computational bits that are the foundation of computers. However, JQI Fellow Christopher Monroe’s group has made progress in working with colleagues at Duke University to make the results reliable, even if quantum computers are sometimes built from the parts that fail. I did. For the first time in an experiment Quantum computing The piece can be better than the worst part used to make it.In a treatise published in a journal *Nature* Today (October 4, 2021), the team shared a way to take this groundbreaking step towards a reliable and practical quantum computer.

In their experiments, researchers combined several qubits (quantum versions of bits) and worked together as a single unit called a logical qubit. They created logical cue bits based on quantum error correction code, so unlike individual physical cue bits, they could easily detect and correct errors, become fault tolerant, contain errors and minimize adverse effects. I was able to suppress it.

“Qubits made up of the same atomic ions are by nature very clean,” said Monroe, a professor at the University of Maryland’s Faculty of Physics and a fellow at the Joint Center for Quantum Information and Computer Science. say. “But at some point, if you need a lot of qubits and manipulations, you need to reduce the errors even further, and it’s easier to add qubits and encode the information differently. Atomic ion error correction code. The advantage of is that it is very efficient and can be turned on flexibly via software control. “

This is the first time that a logical qubit has been shown to be more reliable than the most error-prone steps required to create it. The team was able to successfully start the logical qubit and measure it in 99.4% of the time, but the six quantum operations that are expected to work individually are only about 98.9% of the time.

It may not sound like a big difference, but it’s an important step in building a much larger quantum computer. If each of the six quantum operations is an assembly line worker focusing on one task, the assembly line will only produce the correct initial state for 93.6% of the time (6 times 98.9%). This is an experiment. The improvement is to minimize the chances of quantum errors compounding and ruining the results, just as the imperfections work together in an experiment to allow careful workers to catch each other’s mistakes. ..

Results were achieved using Monroe’s ion trap system at UMD. The system uses up to 32 individual charged atoms (ions) that are laser cooled and suspended on electrodes on the chip. Then, by operating with a laser, each ion is used as a cubit.

“There are 32 laser beams,” says Monroe. “And the atoms are like a line of ducks. Each has its own fully controllable laser beam. I think the atoms form a linear string. I play it like a string. I pick it up with a laser that turns it on and off in a programmable way. And that’s the computer. That’s our central processing device. “

The successful creation of fault-tolerant logical qubits in this system allows researchers to carefully and creatively design and free quantum computing from the unavoidable error constraints of today’s state-of-the-art technology. Showed that there is sex. Fault-tolerant logical qubits are a way to avoid modern qubit errors and can be the basis for reliable, practically large quantum computers.

#### Error correction and failure tolerance

Murphy’s law is unforgiving, so it is important to develop a fault-tolerant qubit that can correct errors. No matter how well you build your machine, you will end up with problems. Bits and qubits may not work on your computer. And the many qubits contained in a practical quantum computer mean that there are many opportunities for errors to creep in.

Fortunately, engineers design their computers to detect errors, such as backing up important information to an additional hard drive or having a second person find a typo before sending it. I can do it. Both people and drives need to ruin their mistakes in order to survive. Redundancy helps ensure final quality, although more work is required to complete the task.

Some popular technologies, such as mobile phones and high-speed modems, are now using error correction to ensure transmission quality and avoid other inconveniences. Error correction using simple redundancy can reduce the chance of uncaught errors unless the procedure is wrong more often than it is correct. For example, if you send or save data three times and trust the majority vote, it’s between 1 / 100th and less than 1/1000.

Therefore, you may never reach perfection, but error correction can improve your computer’s performance as much as you need, as long as you can pay the price of using additional resources. Researchers can use quantum error correction to complement their efforts to create better qubits as well, allowing them to build quantum computers without having to overcome all the errors that quantum devices suffer. I am planning.

“The amazing thing about fault tolerance is the recipe for how to take small, unreliable parts and turn them into highly reliable devices,” said Duke, a professor of electrical and computer engineering and co-author of the treatise. Kenneth Brown says. “And fault-tolerant quantum error correction allows us to create highly reliable quantum computers from defective quantum components.”

However, quantum error correction has its own challenges. Cubits are more complex than traditional bits and can cause problems in many ways. You cannot copy a qubit or check its value in the middle of a calculation. Qubits are advantageous because they can exist in quantum superpositions of multiple states and can be quantum mechanically intertwined with each other. To copy a qubit, you need to know exactly what information is currently stored. Physically, you need to measure the qubit. The measurement then puts it into a single, well-defined quantum state, breaking the superpositions and entanglements in which the quantum computation is built.

Therefore, quantum error correction requires correction of bit errors that are not allowed to be copied. It’s like calibrating while blindfolding. In the mid-1990s, researchers began proposing ways to do this by taking advantage of the subtleties of quantum mechanics, but quantum computers are beginning to be able to experiment with theory.

An important idea is to create a logical qubit from redundant physical qubits so that you can check if the qubits match a particular quantum mechanical fact without knowing their state individually. ..

#### Atoms can’t be improved

Many quantum error correction codes have been proposed to choose from, and some fit more naturally with certain approaches to creating quantum computers. Each method of making a quantum computer has its own type of error and its own strengths. Therefore, to build a practical quantum computer, you need to understand and handle the specific errors and benefits of the approach.

The ion trap-based quantum computer used by Monroe et al. Has the advantage that the individual qubits are identical and very stable. Because qubits are charged ions, each qubit can communicate with everything else in the line via electrical nudge, giving it more freedom compared to systems that require a strong connection to adjacent systems. I have.

“They are atoms of a particular element and isotope, so they are perfectly replicable,” says Monroe. “And if you store the coherence in a cubit and leave them alone, it will exist essentially forever. Therefore, if you leave it alone, the cubit is perfect. To take advantage of that cubit , You have to poke it with a laser, we have to do something with it, we have to hold it atom The presence of electrodes in the vacuum chamber can cause noise in all of these technical stuff and affect the cubits. “

For Monroe’s system, the number one cause of error is the intertwining operation, the creation of a quantum link between two qubits using a laser pulse. Tangle operation is the part required to operate a quantum computer and combine qubits with logical qubits. Therefore, the team cannot expect logical qubits to store information more stably than individual ionic qubits, but fixing the error that occurs when entwining cubits is an important improvement. ..

Researchers chose the bacon-shawl code as a good match for the strengths and weaknesses of the system. For this project, only 15 of the 32 ions the system could support were needed, two ions were not used as cubits, only to even out the spacing between the other ions. .. The code redundantly encoded a single logical qubit with nine qubits and used four additional qubits to determine where the potential error occurred. Using that information, the detected faulty qubit can theoretically be corrected without compromising the “quantum” of the qubit by measuring the state of the individual qubit.

“A key part of quantum error correction is redundancy, so we needed nine qubits to get one logical qubit,” said Laird Egan, a JQI graduate student who is the lead author of this treatise. increase. “But that redundancy can protect one cubit error with the other eight, which helps to find and fix the error.”

The team successfully used the Bacon-Shor code in the ion trap system. The resulting logical qubit required six entanglement operations. The expected error rate for each operation is 0.7% to 1.5%. However, thanks to the careful design of the code, these errors do not combine to result in even higher error rates when the logical qubits are initially prepared using entanglement operations.

The team observed qubit preparation and measurement errors for only 0.6% of the time. This is less than the lowest error expected for any of the individual entanglement operations. The team was then able to move the logical qubit to the second state with only 0.3% error. The team also deliberately introduced the error and demonstrated that it could be detected.

“This is really a demonstration of quantum error correction that improves the performance of the underlying components for the first time,” says Eagan. “And there’s no reason why you can’t do the same when other platforms scale up. The fact that quantum error correction works is actually a proof of concept.”

As the team continues this series of work, it builds a more difficult logic gate from the qubits, performs a complete cycle of error correction that actively fixes the detected errors, and entangles multiple logic qubits. They say they want to achieve similar success by doing so.

“Until this treatise, everyone has focused on creating a single logical qubit,” says Egan. “And now we’ve created one, so it’s like,” What can we do with two because a single logical cubit works? ” “

See also: Laird Egan, Dripto M. Debroy, Crystal Noel, Andrew Risinger, Daiwei Zhu, Debopriyo Biswas, Michael Newman, Muyuan Li, Kenneth R. Brown, Marko Cetina, Christopher, “Error-Correcting Cubit Fault Tolerant Control” Monroe, October 4, 2021 *Nature*..

DOI: 10.1038 / s41586-021-03928-y

Experiments Show Quantum Computers Can Be Better Than the Sum of Their Parts Source link Experiments Show Quantum Computers Can Be Better Than the Sum of Their Parts

This is a syndicated post. Read the original post at Source link .