Soon, quantum errors will no longer have a right to exist. But is it good?
Quantum computers do not yet know how to solve industrial problems better than classical systems, but gradually approach the main milestone: sustainable computing with bug fixing. The new work of Sandia National Laboratories and Quantinum shows how close to this stage the systems on ions have come. In their paper, the researchers described the results of the Helios verification of the commercial quantum computer Quantinum by 98 qubits.
Kubit is a quantum analogue of bit, but it does not work as ordinary zero and one. Due to quantum states, such elements can participate in calculations that are poorly given to classical machines. The problem is that qubits are extremely sensitive to noise. Error in laser adjustment, unnecessary movement of the atom, failure when reading or imperfect operation between qubits quickly spoil the result. Therefore, for quantum computers today, not speed, but reliability is more important.
Helios showed a very high accuracy of basic operations. For actions with one qubit, the reliability reached 99.9975%, for operations with two qubits - 99.921%. In quantum computing, this indicator is called "fidelity" (can be translated as "loyalty"): it shows how close the real operation coincides with the ideal. The higher the value, the fewer errors accumulate during the program.
For Quantinum, these results make Helios the largest and most reliable system of the company at the moment. But the value of work is broader than one record. The US Department of Energy expects to come to fault-tolerant quantum computers – systems where errors can be detected and corrected right during calculations. Without such a layer, even a large quantum processor remains an experimental machine: it quickly accumulates failures and loses a useful result.
Sandia played the role of an independent appraiser in the project. The laboratory has been developing methods for testing quantum computers for many years and is looking for ways to understand where such systems are wrong. For Helios, experts used different tests, including their own techniques. Special attention was paid to intermediate measurements during the calculation. Such operations allow you to read part of the information, without destroying the entire quantum program, and are needed for future error correction.
Helios belongs to the systems on captured ions. In such computers, qubits are implemented on individual atoms held by electromagnetic fields. Lasers control the states of ions and bind them to computation. This approach has a strong side: individual qubits can be very accurate. But when scaling, engineering difficulties arise - you need to manage an increasing number of ions, lasers, optical channels and measurements.
It is here that Sandia and Quantinum associate further progress with integrated photonics. These are miniature chips that conduct light through microscopic optical channels. In quantum computers, on ions, such components can replace part of the bulky optical infrastructure, reduce energy consumption and simplify system management. The more qubits, the more important it becomes not only the quality of the atoms themselves, but also the entire engineering strapping around them.
Sandia and Quantinum’s cooperation has been going on for four years and has been formalized through a joint research and development agreement. The laboratory helps to design and test components that may enter future platforms. The lab assesses the capabilities and risks of quantum technology for the U.S. government, including cryptography, new materials, pharmaceuticals, energy, sensors and secure communications.
Quantum computers are directly related to national security issues. In the future, they can accelerate the modeling of molecules and materials, help in the development of drugs and new energy technologies, and affect cryptography. Therefore, state laboratories not only create their own prototypes, but also check commercial systems to understand the real level of the industry without advertising estimates of manufacturers.
The results of Helios do not mean that a useful universal quantum computer is already ready. 98 qubits with high accuracy is an important step, but for complex tasks will need systems of much larger scale, stable error correction schemes and a reliable software environment. So far, the industry proves that individual operations can be performed more accurately and that control of errors gradually becomes a practical engineering task.
Quantum computers do not yet know how to solve industrial problems better than classical systems, but gradually approach the main milestone: sustainable computing with bug fixing. The new work of Sandia National Laboratories and Quantinum shows how close to this stage the systems on ions have come. In their paper, the researchers described the results of the Helios verification of the commercial quantum computer Quantinum by 98 qubits.
Kubit is a quantum analogue of bit, but it does not work as ordinary zero and one. Due to quantum states, such elements can participate in calculations that are poorly given to classical machines. The problem is that qubits are extremely sensitive to noise. Error in laser adjustment, unnecessary movement of the atom, failure when reading or imperfect operation between qubits quickly spoil the result. Therefore, for quantum computers today, not speed, but reliability is more important.
Helios showed a very high accuracy of basic operations. For actions with one qubit, the reliability reached 99.9975%, for operations with two qubits - 99.921%. In quantum computing, this indicator is called "fidelity" (can be translated as "loyalty"): it shows how close the real operation coincides with the ideal. The higher the value, the fewer errors accumulate during the program.
For Quantinum, these results make Helios the largest and most reliable system of the company at the moment. But the value of work is broader than one record. The US Department of Energy expects to come to fault-tolerant quantum computers – systems where errors can be detected and corrected right during calculations. Without such a layer, even a large quantum processor remains an experimental machine: it quickly accumulates failures and loses a useful result.
Sandia played the role of an independent appraiser in the project. The laboratory has been developing methods for testing quantum computers for many years and is looking for ways to understand where such systems are wrong. For Helios, experts used different tests, including their own techniques. Special attention was paid to intermediate measurements during the calculation. Such operations allow you to read part of the information, without destroying the entire quantum program, and are needed for future error correction.
Helios belongs to the systems on captured ions. In such computers, qubits are implemented on individual atoms held by electromagnetic fields. Lasers control the states of ions and bind them to computation. This approach has a strong side: individual qubits can be very accurate. But when scaling, engineering difficulties arise - you need to manage an increasing number of ions, lasers, optical channels and measurements.
It is here that Sandia and Quantinum associate further progress with integrated photonics. These are miniature chips that conduct light through microscopic optical channels. In quantum computers, on ions, such components can replace part of the bulky optical infrastructure, reduce energy consumption and simplify system management. The more qubits, the more important it becomes not only the quality of the atoms themselves, but also the entire engineering strapping around them.
Sandia and Quantinum’s cooperation has been going on for four years and has been formalized through a joint research and development agreement. The laboratory helps to design and test components that may enter future platforms. The lab assesses the capabilities and risks of quantum technology for the U.S. government, including cryptography, new materials, pharmaceuticals, energy, sensors and secure communications.
Quantum computers are directly related to national security issues. In the future, they can accelerate the modeling of molecules and materials, help in the development of drugs and new energy technologies, and affect cryptography. Therefore, state laboratories not only create their own prototypes, but also check commercial systems to understand the real level of the industry without advertising estimates of manufacturers.
The results of Helios do not mean that a useful universal quantum computer is already ready. 98 qubits with high accuracy is an important step, but for complex tasks will need systems of much larger scale, stable error correction schemes and a reliable software environment. So far, the industry proves that individual operations can be performed more accurately and that control of errors gradually becomes a practical engineering task.
