Optical computation of a spin glass dynamics with tunable complexity.

Area of Science:

• Physics
• Computer Science
• Optics

Background:

• Spin glasses (SGs) are complex systems with applications across multiple scientific domains.
• Studying SG dynamics is computationally intensive, classified as nondeterministic polynomial-time (NP) hard.
• Existing simulation methods face significant computational challenges.

Purpose of the Study:

• To implement an optical simulation of spin glass models.
• To leverage wavefront-shaping devices and light interference for SG dynamics.
• To demonstrate a computationally advantageous approach for SG simulations.

Main Methods:

• Utilized N segments of a wavefront-shaping device as spin variables.
• Employed light interference downstream of scattering material for random couplings (J matrix).
• Measured light intensity on P targets to determine system energy, implementing a Metropolis algorithm.

Main Results:

• Successfully simulated spin glass dynamics, including paramagnetic, ferromagnetic, and SG phases.
• Demonstrated that the transition temperature (Tg) to the glassy phase increases with parameter α.
• Showcased a computational advantage due to simultaneous realization of interaction terms via light interference.

Conclusions:

• The optical spin glass simulation provides a novel and efficient method for studying complex systems.
• The parallel nature of optical measurements offers a speedup compared to in silico simulations.
• This approach opens new avenues for exploring the phase diagram and dynamics of spin glass models.