Leveraging IO Pad Protection Diodes for Recycled IC Detection and Age Estimation Using Polynomial Regression

The presence of counterfeit recycled ICs (CRICs) in the global semiconductor supply chain is a major concern in the present-day world. These CRICs are less reliable and have become a serious threat to the ICs employed in safety–critical systems. Accurate age prediction for integrated circuits (ICs) is crucial for implementing preventative and mitigation strategies to avoid unexpected failures in the field. By precisely estimating the age of an IC, electronic systems can benefit from improved reliability and performance, as maintenance and replacements can be scheduled proactively, and reducing the risk of sudden breakdowns. Furthermore, accurate age prediction plays a vital role in extending the lifespan of electronic devices, which in turn helps to minimize electronic waste. This not only reduces the environmental impact but also supports the broader goal of green computing by promoting more sustainable and resource-efficient technology practices. In this article, we introduce a method for detecting a CRIC and estimating its age by utilizing the existing input-output (IO) pad structures targeting sensorless chips. The proposed methodology estimates age by measuring the voltage drop across the protection diodes present in the IO pad structure and applying this voltage drop to the proposed polynomial regression model. This methodology requires no additional sensory circuit, resulting in no area overhead. As there is no requirement for a special on-chip sensor, the proposed methodology can be used to detect the age of an IC in production. Our proposed polynomial regression model achieves a mean squared error (MSE) of 1.77 h, with a minimum improvement of 99.7% over the state-of-the-art methodologies.