НВ (Новое Время)

Scientists Discover Limitations of Modern Quantum Computing Due to 'Forgetting' Effect

An international team of researchers from leading European universities has uncovered a significant limitation in modern quantum computing, termed the 'forgetting' effect. This discovery, published in the esteemed scientific journal Nature Physics, has profound implications for the future development of quantum technologies.

An international group of researchers, consisting of scientists from leading universities across Europe, has revealed a crucial limitation in contemporary quantum computing, which they have dubbed the 'forgetting' effect. This groundbreaking finding, published in the prestigious scientific journal Nature Physics, carries significant consequences for the ongoing advancement of quantum technologies.

According to the research results, the accumulation of quantum noise imposes a strict limit on the number of operations that a quantum computer can perform sequentially without data loss. This indicates that even in the most advanced quantum systems capable of executing complex calculations, there are constraints associated with the influence of external noise.

The scientists compare the operation of a quantum circuit to a chain of dominoes: ideally, each piece should fall precisely on schedule. However, in reality, qubits, which are the quantum bits, are subjected to constant external interference—noise. Utilizing complex mathematical analysis, the research team was able to track how this noise propagates through the quantum circuit.

The findings of their study revealed that in deeply noisy circuits, only the last few steps truly affect the final outcome. All preceding calculations are effectively erased or rendered insignificant under the pressure of accumulated errors. This discovery is critical for engineering, as it explains why simply increasing the number of steps (circuit depth) does not lead to enhanced performance of quantum computers.

Despite their complexity, modern quantum machines often behave like much simpler 'surface' systems. This revelation also highlights the illusory nature of successes in training noisy circuits: they may appear to be compliant simply because the noise has already reduced their effective complexity to a level that can be replicated even on conventional classical computers.

The authors of the study emphasize that further progress in the field of quantum technologies hinges not on the number of qubits or steps but on a radical reduction in noise levels. Without overcoming this barrier, quantum machines will remain limited in solving problems that are inaccessible to traditional computer systems. This discovery calls into question many of the hopes placed on quantum computing and points to the necessity for new approaches in this area.