Towards the True Switching-ON of Transistors

TL;DR

This paper introduces a first-principles paradigm to fundamentally understand and predict transistor switching behavior, significantly improving accuracy and enabling sustainable development in electrical engineering.

Contribution

The paper presents a novel first-principles framework that explains transistor switching phenomena and improves Eon prediction accuracy by an average of 17 times.

Findings

Eon prediction error reduced to 0.88%-11.60% from 34.41%-80.05%.

Unified physical explanation of switching-ON phenomena across scenarios.

Transformative shift from empirical to first-principles analysis of transistor switching.

Abstract

Transistors are core component across all domains of electrical and electronic engineering (EEE), such as data centers, electrified transportation, robotics, renewables and grid applications, etc. Transistors' switching behavior governs energy loss, carbon emissions, cooling demand, water use, lifetime, material use and cost etc. throughout EEE. Despite near a century since the transistor's invention, the understanding of transistor switching remains fragmented: switching is treated as a black box relying on observed waveforms, cannot be explained using physical laws alone, and is not integrated into circuit theory. This forms one of the most critical barriers to recognizing the true physical boundaries, prohibiting more sustainable solutions. For example, the conventional Eon prediction model, derived from the conventional switching analysis, exhibits significant prediction errors (ranging from 34.41% to 80.05%). Here we present a unified first-principles paradigm to explain the switching phenomena. Using this paradigm, we revealed the physical origins and mechanisms of switching-ON phenomena across scenarios, and derived the proposed Eon prediction model, with error ranging from 0.88% to 11.60%, achieving a 17-fold average improvement. These results demonstrate the unprecedented power of the proposed paradigm: textbook-level foundations are established, transforming the fundamental understanding of transistor switching from empirical to first-principles analysis, and simultaneously stimulating follow-up research and applications for sustainable development across disciplines.