18670512. DATA STORAGE DEVICE WITH NOISE INJECTION simplified abstract (Western Digital Technologies, Inc.)

DATA STORAGE DEVICE WITH NOISE INJECTION

Organization Name

Western Digital Technologies, Inc.

Inventor(s)

Daniel Joseph Linnen of Limestone TN (US)

Kirubakaran Periyannan of Saratoga CA (US)

Ramanathan Muthiah of Bangalore (IN)

DATA STORAGE DEVICE WITH NOISE INJECTION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18670512 titled 'DATA STORAGE DEVICE WITH NOISE INJECTION

Simplified Explanation: The patent application discloses procedures for injecting noise into a non-volatile memory (NVM) array to degrade information intentionally.

Key Features and Innovation:

• Noise injection through adjusting read voltages to induce bit flips
• Iterative combination of random data to achieve a target percentage of random 1s or 0s
• Emulating dead pixels in charge-coupled device (CCD) pixels by zeroing out pixels
• Adjusting timing, voltage, and/or current values to induce bit flips during data transfer
• Use of noise-injected data for dataset augmentation or testing deep neural networks (DNNs)

Potential Applications: The technology can be used for dataset augmentation, testing deep neural networks, and potentially in other applications requiring noise-injected data.

Problems Solved: The technology addresses the need for intentionally degrading information in NVM arrays for various purposes such as testing and augmentation.

Benefits:

• Improved testing of deep neural networks
• Enhanced dataset augmentation capabilities
• Controlled degradation of information for specific applications

Commercial Applications: Potential commercial applications include data testing services for deep neural networks, data augmentation tools for machine learning, and quality control in memory storage devices.

Prior Art: Readers interested in prior art related to noise injection in NVM arrays may explore research on data degradation techniques in memory devices and testing methodologies for deep neural networks.

Frequently Updated Research: Researchers may find updated studies on data degradation techniques, noise injection methods, and their applications in machine learning and memory technologies.

Questions about Noise Injection in NVM Arrays: 1. What are the potential risks associated with injecting noise into non-volatile memory arrays? 2. How does the iterative combination of random data help achieve a target percentage of random 1s or 0s in the injected noise?

Original Abstract Submitted

Noise injection procedures implemented on the die of a non-volatile memory (NVM) array are disclosed. In one example, noise is injected into data by adjusting read voltages to induce bit flips while using feedback to achieve a target amount of information degradation. In another example, random data is iteratively combined with itself to achieve a target percentage of random 1s or 0s, then the random data is combined with data read from the NVM array. In other examples, pixels are randomly zeroed out to emulate dead charge coupled device (CCD) pixels. In still other examples, the timing, voltage, and/or current values used within circuits while transferring data to/from latches or bitlines are adjusted outside their specified margins to induce bit flips to inject noise into the data. The noise-injected data may be used, for example, for dataset augmentation or for the testing of deep neural networks (DNNs).