Despite the long and crucial role of traditional solid-state physics in current silicon-based technologies, next-generation neuromorphic, non-volatile memory, and energy devices—key components in the era of the internet of things (IOT) — require novel working principles that incorporate quantum physics emerging in low-dimensional devices. The main research direction for future devices is to realize ‘ultralow device operation energy’, ‘ultrahigh device operation speed’, and ‘large-scale device integration (up to 10¹⁵)’, which calls for exploring diverse quantum phenomena in low-dimensional device components. In this talk, I will present some of our recent efforts to establish new device physics for energy-intelligent devices, which could be a milestone for the promising future devices. In particular, dynamic convolutional neural networks, phase transition, and other intriguing quantum physics in low-dimensional materials will be discussed along with logic devices, neuromorphic computing, and energy device applications.