Researchers Submit Patent Application, “Health Management for Magnetic Storage Media”, for Approval (USPTO 20230020009): Marvell Asia Pte Ltd.

2023 FEB 02 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- From Washington, D.C., NewsRx journalists report that a patent application by the inventor Nangare, Nitin (Sunnyvale, CA, US), filed on September 19, 2022, was made available online on January 19, 2023.

The patent’s assignee is Marvell Asia Pte Ltd. (Singapore, Singapore).

News editors obtained the following quote from the background information supplied by the inventors: “Electronic devices provide many services to modern society. These services enable an electronic device to provide entertainment, assist with scientific research and development, and provide many modern-day conveniences. Many of these services create or use data, which the electronic device stores. This data may include digital media such as books or movies, algorithms that execute complex simulations, personal user data, applications, and so forth. To avoid exceeding data storage limits, it is beneficial to increase the data storage capacity of the electronic device and avoid deleting data, limiting services, or purchasing additional external storage devices.

“Many electronic devices use media drives to store data on disks, such as a hard-disk drive (HDD). Generally, data stored to a HDD is accessed by moving a read element or write element across a surface of a media disk as the disk rotates at high speed. As such, HDDs are a complex system of interlinked mechanical, electrical, and magnetic parts that operate in concert to facilitate the storage and access of data. Outside of catastrophic events, such as being dropped, these mechanical, electrical, and magnetic parts of the HDD wear gradually during the operating life of the HDD until a critical failure of one of the parts prevents data access. Without knowing how much of this wear has occurred over time, a HDD may fail unexpectedly before replacement or data backup, resulting in the loss of data or reduced uptime of data storage services.”

As a supplement to the background information on this patent application, NewsRx correspondents also obtained the inventor’s summary information for this patent application: “This summary is provided to introduce subject matter that is further described in the Detailed Description and Drawings. Accordingly, this Summary should not be considered to describe essential features nor used to limit the scope of the claimed subject matter.

“In some aspects, a media health manager of magnetic storage media implements a method that receives a request to read a sector of magnetic storage media that resides in a zone of magnetic storage media. The method determines, with a read channel, read metrics for the sector of the magnetic storage media that is read based on the request. The method includes accessing read metrics of the zone of magnetic storage media and updating the read metrics of the zone based on the read metrics determined for the sector to provide updated read metrics for the zone of magnetic storage media. The method then determines, with a neural network, a health score for the zone of magnetic storage media based on the updated read metrics of the zone of magnetic storage media.

“In other aspects, an apparatus comprises an interface to communicate data with a host, a disk of magnetic storage media to store the data, a read channel to read the data from sectors of the magnetic storage media, and a neural network to determine media health scores. The apparatus also includes a media health manager configured to determine, with the read channel, read metrics for a sector of the magnetic storage media from which the data is read. The media health manager may also determine, from multiple zones of magnetic storage media, a zone of magnetic storage media in which the sector resides. Each zone of magnetic storage media may correspond to a respective subset of tracks on the disk of magnetic storage media. The media health manager accesses read metrics of the zone of magnetic storage media and updates the read metrics of the zone based on the read metrics determined for the sector to provide updated read metrics for the zone of magnetic storage media. The media health manager then determines, with the neural network, a health score for the zone of magnetic storage media based on the updated read metrics of the zone of magnetic storage media.

“In yet other aspects, a System-on-Chip (SoC) is described that includes a pre-amplifier interface to receive signals corresponding to data read from sectors of magnetic storage media. The SoC also includes read channel logic to process the signals corresponding to the data read from the sectors of magnetic storage media, a hardware-based processor, and a memory storing processor-executable instructions to implement a media health manager. The media health manager determines, with the read channel logic, read metrics for a sector of the magnetic storage media from which the data is read. The media health manager may also determine, from multiple zones of magnetic storage media, a zone of magnetic storage media in which the sector resides. Each zone of magnetic storage media may correspond to a respective subset of tracks on a disk of magnetic storage media. The media health manager accesses read metrics of the zone of magnetic storage media and updates the read metrics of the zone based on the read metrics determined for the sector to provide updated read metrics for the zone of magnetic storage media. The media health manager then determines, with the neural network, a health score for the zone of magnetic storage media based on the updated read metrics of the zone of magnetic storage media.

“The details of one or more implementations are set forth in the accompanying drawings and the following description. Other features and advantages will be apparent from the description and drawings, and from the claims.”

The claims supplied by the inventors are:

“1. A method for managing health of magnetic storage media of a storage media drive, comprising: receiving a request to read a sector of the magnetic storage media that resides in a zone of the magnetic storage media; determining, with a read channel of the storage media drive, a read metric for the sector of the magnetic storage media that is read based on the request; accessing read metrics of the zone of the magnetic storage media, the read metrics of the zone including respective read metrics for the sector that is read and others of the multiple sectors of the zone of the magnetic storage media; updating the read metrics of the zone of the magnetic storage media based on the read metric determined for the sector of the magnetic storage media to provide updated read metrics for the zone of the magnetic storage media; and determining, with a neural network, a health score for the zone of the magnetic storage media in which the sector resides based on the updated read metrics of the zone of the magnetic storage media.

“2. The method of claim 1, wherein the neural network is configured to map the read metrics of the read channel to one of multiple health scores to determine the health score for the zone of the magnetic storage media.

“3. The method of claim 1, wherein: the neural network comprises a pre-trained neural network configured to provide at least one of magnetic media health scoring, component wear estimation, or storage media drive health prediction; or the health score for the zone of the magnetic storage media is indicative of a level of wear, a level of degradation, or a level of reliability of the zone of the magnetic storage media.

“4. The method of claim 3, wherein the pre-trained neural network is trained through supervised learning, and the pre-trained neural network is configured to implement unsupervised learning or reinforcement learning based on at least the updated read metrics for the zone of magnetic storage media.

“5. The method of claim 1, wherein the neural network comprises at least an input layer, an output layer, and one or more hidden intermediate layers positioned between the input layer and the output layers of the neural network.

“6. The method of claim 1, wherein the neural network comprises one of a deep neural network, a convolutional neural network, or a recurrent neural network.

“7. The method of claim 1, wherein: the neural network comprises or is configured to implement a regression model to determine the health score for the zone of the magnetic storage media; and the regression model of the neural network or implemented by the neural network comprises a single linear regression model, multiple linear regression models, a logistical regression model, a stepwise regression model, a multi-variate adaptive regression model, or a locally estimated scatterplot model.

“8. The method of claim 1, further comprising: storing the health score of the zone of the magnetic storage media to a table of zone information; or transmitting the health score of the zone of the magnetic storage media to a storage media controller of the storage media drive.

“9. An apparatus comprising: an interface to communicate data with a host; a disk of magnetic storage media to store the data; a read channel to read the data from sectors of the magnetic storage media; a neural network configured to determine media health scores; and a media health manager configured to: determine, with the read channel, read metrics for a sector of the magnetic storage media from which the data is read; determine, from multiple zones of the magnetic storage media, a zone of the magnetic storage media in which the sector resides; access read metrics of the zone of the magnetic storage media, the read metrics of the zone of the magnetic storage media including respective read metrics for the sector from which the data is read and multiple others of the sectors of the zone of the magnetic storage media; update the read metrics of the zone of the magnetic storage media based on the read metrics determined for the sector of the magnetic storage media to provide updated read metrics for the zone of the magnetic storage media; and determine, with the neural network, a health score for the zone of the magnetic storage media in which the sector resides based on the updated read metrics of the zone of the magnetic storage media.

“10. The apparatus of claim 9, wherein the neural network comprises a pre-trained neural network configured to provide at least one of magnetic media health scoring, component wear estimation, or storage media drive health prediction.

“11. The apparatus of claim 10, wherein the pre-trained neural network is trained through supervised learning, and the pre-trained neural network is configured to implement unsupervised learning or reinforcement learning based on at least the updated read metrics for the zone of magnetic storage media.

“12. The apparatus of claim 9, wherein the neural network comprises one of a deep neural network, a convolutional neural network, or a recurrent neural network.

“13. The apparatus of claim 9, wherein the neural network comprises or is configured to implement a regression model to determine the health score for the zone of the magnetic storage media.

“14. The apparatus of claim 9, wherein the regression model of or implemented by the neural network comprises a single linear regression model, multiple linear regression models, a logistical regression model, a stepwise regression model, a multi-variate adaptive regression model, or a locally estimated scatterplot model.

“15. The apparatus of claim 10, wherein the read metrics for the sector of the magnetic storage media include one or more of an unrecovered sector rate, an average retry method rate, an average of off-track detections, an average number of seek errors, an average number of errors corrected by error-correcting code (ECC), an average syndrome weight in ECC, an average number of ECC iterations, an average of log likelihood ratio (LLR) probability density function (PDF) statistics, an average mean square error (MSE) of front-end loop signals, an average length of detected media defects, an average length of analog-to-digital converter (ADC) saturation, an average delta of an adaptive parameter, or an average of sync mark errors.

“16. A System-on-Chip (SoC) comprising: a pre-amplifier interface to receive signals corresponding to data read from sectors of magnetic storage media; read channel logic to process the signals corresponding to the data read from the sectors of the magnetic storage media; a hardware-based processor; a memory storing processor-executable instructions that, responsive to execution by the hardware-based processor, implement a media health manager to: determine, with the read channel logic, read metrics for a sector of the magnetic storage media from which the data is read; determine, from multiple zones of the magnetic storage media, a zone of magnetic storage media in which the sector resides; access read metrics of the zone of the magnetic storage media, the read metrics of the zone of the magnetic storage media including respective read metrics for the sector from which the data is read and multiple others of the sectors of the zone of the magnetic storage media; update the read metrics of the zone of the magnetic storage media based on the read metrics determined for the sector of the magnetic storage media to provide updated read metrics for the zone of the magnetic storage media; and determine, with a neural network, a health score for the zone of the magnetic storage media in which the sector resides based on the updated read metrics of the zone of the magnetic storage media.

“17. The SoC of claim 16, wherein: the neural network comprises a pre-trained neural network configured to provide at least one of magnetic media health scoring, component wear estimation, or storage media drive health prediction; or the health score for the zone of the magnetic storage media is indicative of a level of wear, a level of degradation, or a level of reliability of the zone of the magnetic storage media.

“18. The SoC of claim 17, wherein the pre-trained neural network is trained through supervised learning, and the pre-trained neural network is configured to implement unsupervised learning or reinforcement learning based on at least the updated read metrics for the zone of magnetic storage media.

“19. The SoC of claim 16, wherein the neural network comprises one of a deep neural network, a convolutional neural network, or a recurrent neural network.

“20. The SoC of claim 16, wherein: the neural network comprises or is configured to implement a regression model to determine the health score for the zone of the magnetic storage media; and the regression model of or implemented by the neural network comprises a single linear regression model, multiple linear regression models, a logistical regression model, a stepwise regression model, a multi-variate adaptive regression model, or a locally estimated scatterplot model.”

For additional information on this patent application, see: Nangare, Nitin. Health Management for Magnetic Storage Media. U.S. Patent Application Number 20230020009, filed September 19, 2022 and posted January 19, 2023. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(20230020009)&db=US-PGPUB&type=ids

(Our reports deliver fact-based news of research and discoveries from around the world.)