As conventional computing approaches its physical limits, probabilistic computing has emerged as a promising paradigm for solving complex, combinational-optimization problems beyond the scope of traditional digital logic. This talk explores how probabilistic computing not only offers algorithmic advantages in areas such as optimization and machine learning, but also opens a new platform for the design and utilization of quantum materials and superconducting devices. By leveraging intrinsic stochasticity mainly from thermal noise and fluctuations—probabilistic hardware can enable inherently parallel and energy-efficient computation. In this talk, I will present the original idea of probabilistic computing, its computational architecture (Boltzmann Machine), and recent progress in the development of probabilistic bits (p-bits) implemented with spintronics devices and superconducting Josephson junctions. This interdisciplinary framework unites device physics, materials science, and unconventional computing architectures, paving the way toward scalable, hybrid platforms for intelligent and energy-efficient computation.