No quantum advantage (yet) in the world of tensor networks
Classical computation is more accurate that quantum annealer for some Ising spin glasses The post No quantum advantage (yet) in the world of tensor networks appeared first on Physics World .
Researchers at the Flatiron Institute in New York City recently employed a novel tensor network scheme to simulate the evolution of Ising spin glasses over time. Their classical method performed more accurately than the latest quantum annealers executing the same task. This breakthrough expands the gap between classical and quantum computers in simulating many-particle quantum systems.
Quantum systems, characterized by exponential complexity with an increase in particles, provide an ideal benchmark for evaluating computational techniques. In a classical computer, each particle's spin is encoded by a bit, either zero or one, while a quantum computer employs qubits, which can exist in a superposition of states. However, a recent study by researchers from the Flatiron Institute found that a classical tensor network approach could perform just as well as, if not better than, the quantum annealer in simulating Ising spin glass dynamics.
The researchers used tensor networks, akin to LEGO bricks, to represent the interactions between particles in a many-particle system. By linking tensors together through a process called contraction, the network can more accurately encode the correlations between particles. To manage the increasing complexity as the tensor network evolves in time, the team used a belief propagation approach, which allows the computation to keep pace with the entanglement inherent in the system's time evolution.
The findings, published in Science, highlight the rapid progress being made in classical simulation of quantum physics and the ongoing competition between classical and quantum computation.
Written by urgent.news from Physics World's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
