At present, two experimental approaches for the realization of qubits are considered to be the most advanced: superconducting circuits and trapped ions. The former store quantum information in electronic components, the latter at different energy levels of individual atoms. In superconducting circuits, it was recently shown experimentally for the first time that quantum computers can perform highly specialized tasks that classical computers fail to do. The ion-based method, on the other hand, is characterized by the fact that the error rate of computing operations has always been much lower than with any other approach.
The ion method now developed by scientists from the University of Hannvover and the German national metrology agency PTB further reduces the error rate and thus delivers reliable calculation results much faster. It follows an approach in which the ions are held in a vacuum above a chip structure by means of electric fields.
The computing operations on the qubits are performed by sending microwave signals through special conductor loops embedded in the chip structure. Usually, extremely precisely controlled laser beams are used to perform computing operations. The use of microwaves has the advantage that microwave technology is very advanced and in widespread use which makes it relatively cheap to use. And that it is comparatively easy to control these fields.
The researchers have investigated how to perform the computing operations on the qubits most efficiently. This is a question that is also of great relevance in today’s computer chips, because in the end the energy required per computing operation decides how many of them can be performed per second before the chip gets too hot. In the case of the ion-microwave quantum computer, the researchers were able to show that specially shaped microwave pulses, in which the microwave field is slowly built up and then broken down again, have error rates 100 times lower than a calculation operation in which the fields are simply switched on and off for the same energy input, despite the presence of interference sources.
To this end, the team had introduced additional, precisely controlled sources of interference into the experiment and determined the calculation errors for sources of interference of varying intensity and for both pulse forms. “For our experiment this made a huge difference,” says Giorgio Zarantonello, one of the authors of the study. “In the past, for good arithmetic operations, we had to try and optimize for a long time until we caught a moment when the sources of interference were very small. Now we can simply switch on our experiment and it works!”.
Now that the scientists have been able to show that elementary arithmetic operations can be realized with low error rates, they want to achieve the same for more complex tasks. Their goal is to achieve significantly less than one error every ten thousand operations. Only then does it make sense to extend the application to many qubits. To this end, the scientists have already developed a patented manufacturing process that makes it possible to store and manipulate many qubits in a chip structure.
The work was supported by the Quantum Technology Flagship Project of the EU, among others. Within the next decade, the EU intends to invest 1 billion euros to make findings from quantum physical basic research technologically usable. Researchers from Hanover and Braunschweig are working here within the “MicroQC” project together with colleagues from Siegen, Sussex, Jerusalem and Sofia.
Original publication: https://doi.org/10.1103/PhysRevLett.123.260503