Samsung Patent: Machine Learning for Storage Device Reliability

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted Samsung Electronics a patent for a method using machine learning to predict the reliability of storage devices. The patent, filed in 2021, outlines a system that selects optimal machine learning models based on device deterioration information to generate reliability predictions.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has granted Samsung Electronics a patent (US12586004B2) for a novel method of predicting the reliability of storage devices, particularly those with nonvolatile memories. The patented technology utilizes machine learning models, selected based on device deterioration characteristics and phase information, to forecast reliability. This prediction then informs how the storage device operates, including read operations.

This patent grant is primarily informational for manufacturers and technology developers in the storage sector. While it does not impose new regulatory obligations or compliance deadlines, it signifies a technological advancement in predictive maintenance for storage hardware. Companies involved in developing or utilizing advanced storage solutions may wish to review the patent for competitive intelligence and potential licensing opportunities.

Source document (simplified)

← USPTO Patent Grants

Methods of predicting reliability information of storage devices and methods of operating storage devices

Grant US12586004B2 Kind: B2 Mar 24, 2026

Assignee

Samsung Electronics Co., Ltd.

Inventors

Chanha Kim, Youngeun Kim, Kangho Roh, Younghoon Jung, Mijung Cho

Abstract

In a method of operating a storage device including a plurality of nonvolatile memories, reliability information of the storage device is predicted. A read operation on the storage device is performed based on a result of predicting the reliability information. In the predicting the reliability information of the storage device, a model request signal is outputted by selecting one of a plurality of machine learning models as an optimal machine learning model based on deterioration characteristic information and deterioration phase information. The model request signal corresponds to the optimal machine learning model. The plurality of machine learning models are used to generate first reliability information related to the plurality of nonvolatile memories. First parameters of the optimal machine learning model may be received based on the model request signal. The first reliability information is generated based on the deterioration characteristic information and the first parameters.