SRAM PUF Startup Testing for Process Variation Quantification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor manufacturing, process variations during CMOS technology scaling down lead to transistor characteristic inconsistencies, resulting in performance losses, increased static currents, and memory read/write failures, necessitating conservative circuit design with higher costs and power consumption.
Innovation Solution
A method to measure and quantify process-induced variability using a physically unclonable function (PUF) device with an SRAM array, where a controller iteratively powers up PUF cells to detect state changes and evaluate process variations by comparing current and previous states, determining the reliability of bits for signature generation without additional hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative circuit design is used to cope with process variations, then circuit reliability is improved, but area and power consumption increase
Solution Approach 1:
The patent changes the parameter of design margin by introducing process variation distribution measurement. Instead of using fixed conservative margins, the system dynamically adjusts design parameters based on measured process variation characteristics, allowing optimization between reliability and area/power consumption.
2Reliability
If conservative circuit design is used to cope with process variations, then circuit reliability is improved, but power consumption increases
Solution Approach 1:
The system dynamically changes power consumption parameters based on measured process variation distribution. By understanding the actual variation characteristics, the design can optimize power settings rather than using fixed conservative values, reducing unnecessary power consumption while maintaining reliability.
3Manufacturing precision
If process variation measurement and quantification is implemented, then process distribution tightness is improved, but measurement and evaluation complexity increases
Solution Approach 1:
The patent implements self-service measurement where the PUF device itself is used as the measurement tool. The PUF cells are powered up and their state changes are detected to directly measure process variation distribution, eliminating the need for external complex measurement equipment and simplifying the measurement process.
Data Source
AI summary
A process variation evaluation method and a process variation evaluating device are provided. The process variation evaluation method is adapted to an electronic device having a PUF device, a recording array and a controller, and includes: iteratively powering up the PUF device in various conditions and detecting a state change of each of the PUF cells in the PUF device by the controller, in which the controller compares a current state in a current iteration and a previous state in a previous iteration of each PUF cell to detect the state change and setting a value indicating the state change in a corresponding bit in the recording array; counting a number of bits with the set value in the recording array after a plurality of iterations; and evaluating a variation of a process according to the counted number by the controller.