Aging Analysis of Various SRAM Cells Under Diverse Workload Condition

SRAM cells plays pivotal role in Internet of Things (IoT), and in-memory computing (IMC) systems, where stability and reliability are critical for future designs. SRAM stability, governed by the read ratio (RR) and cell ratio (CR), is compromised by device reliability challenges such as bias temperature instability (BTI), process variations, and hot carrier injection (HCI), which degrade noise margins and operational speed. While modern systems prioritize energy efficiency, reliability analysis often remains underemphasized. This work investigates SRAM cell performance under diverse workload conditions, categorizing cells into high-density (HD) and high-speed (HS) configurations by adjusting transistor sizes. For 6T cells, CR is varied, whereas 8T and 10T cells focus on RR adjustments at fixed CR values. Stability metrics, including read static noise margin (RSNM) and write noise margin (WNM), are quantified via the Butterfly method. The aging analysis highlights trade-offs between density, speed, and reliability, offering insights for optimizing SRAM design in reliability-sensitive applications.