Beyond Traditional Computing Architecture: the Evolution of All-in-One Devices

Traditional computing architectures face fundamental limitations due to the physical separation between sensing, storage, and computing units, resulting in data transmission bottlenecks and inefficient energy consumption from processing redundant information. This review examines the emerging field of all-in-one devices that integrate sensing, memory, and computing functions into a single device, offering a promising solution to these challenges. By enabling computation within the sensing system itself, these devices can significantly improve data processing speed and energy efficiency. The review covers recent developments across multiple sensing domains, including optical, pressure, gas, and auditory sensors. We analyze how this architectural paradigm shift enables applications requiring low latency and multi-modal sensing capabilities, from autonomous vehicles to wearable medical devices. The integration of multiple functionalities in all-in-one devices represents a significant step toward more efficient and responsive computing systems for the growing demands of intelligent systems.