Stacked-Die TRNG Architecture for Small-Footprint Secure Randomness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
True random number generators (TRNGs) integrated in semiconductor devices face challenges in reducing the integrated circuit footprint while maintaining security, as existing solutions using ring oscillators are vulnerable to external attacks and costly in terms of footprint.
Innovation Solution
A semiconductor device design that separates the randomness source circuit and the randomness harvesting circuit across two dies, with the randomness source circuit integrated in the back end of line, allowing for a smaller footprint and improved security by using selector self-oscillation or non-volatile memory circuits to generate oscillation signals, which are then converted into random bits by the harvesting circuit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If ring oscillator TRNG is integrated in the front end of line, then randomness generation is achieved, but footprint area increases
Solution Approach 1:
The patent divides the TRNG into two separate circuits: a randomness source circuit (RSC) that generates oscillation signals and a randomness harvesting circuit (RHC) that converts them to random bits. This segmentation allows the RSC to be placed in the back end of line using existing memory structures, reducing the footprint in the front end while maintaining randomness generation capability.
Solution Approach 2:
The patent extracts the randomness source function from traditional front-end oscillator circuits and relocates it to the back end of line, utilizing existing memory array structures. This extraction reduces the footprint area in critical front-end regions while preserving the TRNG functionality through the separated RSC-RHC architecture.
2Area of stationary object
If ring oscillator TRNG is integrated in external circuit, then footprint area is reduced, but security vulnerability increases
Solution Approach 1:
By segmenting the TRNG into RSC and RHC placed on different dies or locations, the patent reduces security vulnerabilities. The RSC in the back end uses secure memory structures, while the RHC processes the signals. This physical and functional separation protects against external attacks while maintaining compact footprint.
Solution Approach 2:
The patent introduces an intermediary connection between the RSC and RHC that allows secure transmission of oscillation signals. This intermediary interface is designed to protect against external attacks while enabling the footprint reduction benefits of external integration.
3Reliability
If traditional TRNG integration is used, then randomness generation is achieved, but power consumption increases
Solution Approach 1:
The RSC utilizes existing memory array structures and their inherent physical phenomena (such as threshold voltage variations and leakage currents) to generate oscillation signals without requiring additional dedicated oscillation circuitry. This self-service approach reduces power consumption while maintaining randomness generation capability.
Solution Approach 2:
The patent makes the memory array structures serve dual purposes: storing data and generating random oscillation signals. This multi-functionality eliminates the need for separate oscillator circuits, thereby reducing overall power consumption while achieving reliable randomness generation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This design reduces the footprint of the true random number generator, enhances security by providing asynchronous randomness generation, and lowers power consumption while maintaining high randomness quality, as validated by SP800-22 statistical tests.
Implementation Method 1
a selector self-oscillation circuit, a selector-controlled oscillation circuit
Implementation Method 2
based on a physical phenomenon that is expected to be random... thermal noise, radioactive decay, shot noise, avalanche noise
Implementation Method 3
The randomness harvesting circuit may include an edge detection circuit that detects switching events of the oscillation signal
Data Source
AI summary
The present disclosure provides a semiconductor device, which includes a first die and a second die. The first die includes a randomness harvesting circuit. The second die includes a memory array, and the second die is vertically stacked on the first die. The memory array includes a randomness source circuit, and a true random number is generated using the randomness source circuit on the second die and the randomness harvesting circuit on the first die.