⇧ [VIDÉO] You may also like this partner content (after the ad)
Recently, IBM has already made headlines by unveiling its latest quantum computer prototype. Capable of using 433 qubits. However, the company made several revelations during its IBM Quantum Summit. He says he baked a 127-qubit circuit that demonstrated its ability to run a large computer program.
Why should it be surprising that an IBM machine can run a program? After all, you might think that this is the most logical use for a computer. However, in the field of quantum computing, there are still many limitations that are difficult to overcome.
Running the program is one of them. In other words, the more research groups can build a quantum computer, the more difficult it is to put them to a concrete application. One of the major obstacles to their implementation is the complexity of correcting computational errors. In fact, the purpose of a quantum computer is to perform calculations based on the principles of quantum physics. In short, quantum physics, a highly complex field, is the science dealing with the behavior of matter and light at an atomic level. By studying how matter behaves at this level, some physical principles have been revealed that are very different from what we know.
Among them is “quantum superposition”, which is crucial to the operation of quantum computers. Roughly speaking, quantum superposition can be summed up as the fact that something can be “in two states at once”. For a classical computer, the basic unit of information is the “bit”. It can be either “0” or “1”. In a quantum computer, things work a little differently. However, there is a kind of equivalent called “Cubit”. Thanks to the famous law of superposition, qubits can somehow be both 0 and 1, and even positions in between, such as 01, 10, 11… which allows us to use discrete computing power.
“Accept” errors and correct them better
However, this ability to account for common miscalculations is a weak point in quantum computing. ” Quantum computers are inherently more sensitive to perturbations “, thus confirmed A statement from the University of Innsbruck, discusses the error correction solution. These popular errors can make it very difficult to run a program efficiently.
IBM researchers were able to use a method that allows them to fix a backend. The quantum circuit they used was made up of 127 qubits, which performed more than 1700 unique operations. The operations they implemented on the computer aimed to calculate what happens when a chain of electron-like particles is suddenly forced to interact with each other. The choice to solve this problem is not trivial. Indeed, conventional computers can partially solve these questions, but they can only make approximations.
Ironically, the method they used to eliminate mistakes was… repeating them. To understand how they affect the outcome, they make sure to duplicate and multiply errors that are already uncorrectable. By doing so, they were able to mathematically correct the errors to get the correct result.