High-performance hardware primitives based on sub-10 nm nanodiodes for cryptography applications

Abstract

In the post-silicon era, true random number generators (TRNGs) based on various memristors or nanomaterials are garnering significant interest across diverse disciplines, emerging as a research hotspot in the field of hardware security. In particular, high-entropy of TRNGs is consistently pursued as a key indicator to generate high-security random bit streams. However, rarely any candidate can meet the high-entropy demand while comprehensively considering practical feasibility, such as miniaturization, power consumption, and environmental stability. Here, we demonstrate a compact all-electric quantum TRNG based on a sub-10 nm air-channel nanodiode, displaying near-ideal unpredictability, unrepeatability, and unbiasedness. Notably, this TRNG achieves a record-high min-entropy (>0.99 bits per bit) while maintaining an ultra-high extraction rate of 80% by fully leveraging the information content of original signals, passing all standard randomness tests and exhibiting resilience against machine learning attacks. The high performance is mainly attributed to the intrinsic stochastic behavior of ultrafast tunneling emission–current fluctuations with unpredictable features. In particular, this TRNG also offers full advantages in sub-10 nm scale miniaturization, CMOS-compatible integration, efficient energy consumption (∼1 × 10⁻³ pJ per bit), ultrafast response speed (∼1 ps), and excellent environmental tolerance, making it a promising cryptographic primitive for next-generation nanoelectronics.