Name
Abstract | ||
|---|---|---|
On-chip thermal sensors are essential devices to run modern complicated semiconductor devices safely. The on-chip temperature sensors should be accurate and efficient since we need a number of on-chip temperature sensors to cover the chip-level thermal management. Recently, there have been several attempts to utilize the temperature-dependent properties of memristor devices for temperature sensing. An approach tries to measure the change of resistance after the fixed duration of a write current pulse, and the other attempts to measure the write time variation to complete the memristor state change. In this paper, we propose a statistical method to estimate the temperature without voltage or period measurement. Instead of measuring the write time of the single cell, the proposed method counts the intermediate states of the multiple memristor cells after a certain period of the write accesses where the memristor cells preserve its states with the non-volatility. And then, it quantifies the impact of temperature on the state of the cell array considering a process variation. It provides a completely digital way for temperature estimation without any measurement circuit and analog-to-digital conversion. We investigate the feasibility of the proposed method and tackle some expected design issues. The analysis results show that the proposed statistical temperature sensing method may significantly reduce the power consumption with reasonable timing overhead. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1109/NVMSA.2018.00024 | 2018 IEEE 7th Non-Volatile Memory Systems and Applications Symposium (NVMSA) |
Keywords | Field | DocType |
memristor, process variation, on-chip temperature sensor, statistical sensing, fractional write | Memristor,Computer science,Voltage,Electronic engineering,Process variation,Semiconductor device,Temperature measurement,Analog to digital conversion,Power consumption,Cellular array | Conference |
ISSN | ISBN | Citations |
2575-2561 | 978-1-5386-7404-8 | 0 |
PageRank | References | Authors |
0.34 | 5 | 2 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| The-Nghia Nguyen | 1 | 0 | 0.34 |
| Donghwa Shin | 2 | 396 | 32.34 |