Researchers have developed a quantum chemistry simulation benchmark to evaluate the performance of quantum devices and to control the development of applications for future quantum computers.
Quantum computers use the laws of mechanics and quantum units, known as cubes, to significantly increase the threshold at which information can be sent and processed.
While conventional bits are 0 or 1, the cube is encoded with 0 and 1 or a combination thereof, which allows a large number of data storage options.
Although quantum systems are still at an early stage, they have the potential to be exponentially more powerful than today’s leading classical computer systems and they promise to revolutionize research in the fields of materials, chemistry, high energy physics and the scientific spectrum.
However, because this system is still in its infancy, understanding which applications are suitable for its unique architecture is considered an important field of research.
The researchers calculated the energy states of binding alkaline hydride molecules on the 20-cubic IBM Tokyo and 16-cubic Rigetti Aspen processors. These molecules are simple and their energy is well understood so they can effectively test the work of quantum computers.
By building a quantum computer as a function of several parameters, the team calculated the bonding conditions of these molecules with chemical precision obtained by simulations on a classical computer.
Systematic errors occur when the “noise” inherent in current quantum architecture affects their performance. Because quantum computers are very sensitive (for example, the cube used by the ORNL team is stored in a dilution refrigerator of about 20 milli-degrees (or more than -450 degrees Fahrenheit)), the temperature and vibration of their environment can cause instability, which increases their accuracy.
For example, the noise can cause the cube to rotate 21 degrees, not the desired 20 degrees, which significantly influences the calculation results.
This new indicator characterizes “mixed states” or how media and machines interact very well, the researchers said.
This work is an important step towards universal performance benchmarks that resemble the LINPACK metric used to evaluate the world’s fastest classic computer.
Such leadership requires a system like Summit to enable stable change from devices such as the ORNL team to a larger scale system that is exponentially stronger than anything that works at the moment.
This project will help DOE better understand what will and will not succeed if they advance their mission to realize the potential of quantum computers to overcome the biggest challenges in science and national security, researchers say. The team then plans to calculate exponentially more excited excited states of these molecules to develop new error correction schemes and take the possibility of practical quantum calculations one step closer to reality.