Kabushiki Kaisha Toshiba

Toshiba Electronic Devices & Storage Corporation

Eric R Dunn

Richard M Ehrlich

Thorsten Schmidt

Fernando Anibal Zayas

Mine Wonkyung Budiman

Gary W Calfee

Annie Mylang Le

Hiroshi Sukegawa

Kazuya Takada

Shoichi Aoki

Yosuke Kondo

Kana Furuhashi

Andre C Hall

Daniel Tchen

Motor transportation fixture

Systems and methods for scheduling the execution of disk access commands in a split-actuator hard disk drive are provided. In some embodiments, while a first actuator of the split actuator is in the process of performing a first disk access command (a victim operation), a second disk access command (an aggressor operation) is selected for and executed by a second actuator of the split actuator. The aggressor operation is selected from a queue of disk access commands for the second actuator, and is selected based on being the disk access command in the queue that can be initiated sooner than any other disk access command in the queue without disturbing the victim operation.

Seek scheduling in a split actuator drive

According to an embodiment, each of a plurality of controller chips included in a magnetic disk device includes a buffer control circuit and an arbitration circuit, and controls a corresponding one of a plurality of actuator systems. The first controller chip is connected to a buffer memory via the buffer control circuit included in the first controller chip, and is connected to the second controller chip. The second controller chip is connected to the first controller chip and the third controller chip. The arbitration circuit included in the second controller chip performs arbitration between data transfer between the third controller chip and the first controller chip and data transfer between the first controller chip and an actuator system controlled by the second controller chip among the plurality of actuator systems.

Magnetic disk device

In a magnetic recording drive that includes a shingled magnetic recording (SMR) region and a conventional magnetic recording (CMR) region, the quantity of validation data that is stored by the CMR region is reduced. In the magnetic recording drive, verification data stored in the CMR region is invalidated under certain circumstances, including when an SMR band is closed, when an SMR band is indicated by a host to be reused, when an SMR band is indicated to be finished, and when a last data track of an SMR has data stored therein.

Handling of verification data in disk drive cache

A method for writing data in a disk drive having actuators each controlling arms extending over disk surfaces, including: receiving a write command from a host; receiving from the host data; dividing the data into data blocks; determining: a first surface from the disk surfaces where data is written by a first head of an arm controlled by a first actuator of the actuators; and a second surface from the disk surfaces where data is written by a second head of an arm controlled by a second actuator of the actuators; determining storage blocks of each of the first and the second surface; and writing first data blocks of the divided data blocks to the determined storage blocks of the first surface using the first head while writing second data blocks of the divided data blocks to the determined storage blocks of the second surface using the second head.

Data-center drive with split-actuator that increases read/write performance via data striping

A shingled magnetic recording (SMR) hard disk drive (HDD) is configured with a multi-level cache. To expedite execution of read commands, the SMR HDD is configured to generate and store a Bloom filter in a memory that can be quickly accessed by the drive controller whenever data are stored in certain levels of the multi-level cache. When data are flushed from one level of media cache to an SMR band included in a lower level of media cache, a Bloom filter is generated based on the logical block addresses (LBAs) stored in that SMR band. Thus, when the SMR HDD receives a read command for data that are associated with a particular LBA and are stored in an SMR region of the HDD, the drive controller can query the Bloom filter for each different SMR region of the HDD in which data for that LBA can possibly be stored.

SMR drive with multi-level write-cache for high random-write performance

A method and a system are provided for improving performance of a hybrid drive including a non-volatile semiconductor memory device partitioned into blocks, each of the blocks containing a plurality of sectors, and a magnetic storage device. Performance of the hybrid drive is improved by tracking data types of each sector stored in the blocks, the data types including a first data type, which is data that is unconditionally available for host accesses, a second data type, which is data that is conditionally available for host accesses, and a third data type, which is data unavailable for host accesses, and collecting erasable blocks from the blocks of the non-volatile semiconductor memory device according to the data types. The erasable blocks include a block that contains data of the second data type, such that the host may access from this block even though this block is erasable.

Hybrid-HDD that gives up old NAND data at the last moment

Data is stored in a hybrid HDD that includes a magnetic storage medium and a nonvolatile solid-state device according to multiple modes of operation: a full caching mode, a transitional caching mode, and an HDD only mode. The mode of operation may be selected based on the current condition or performance of individual storage regions in a nonvolatile solid-state device of the hybrid HDD, or on the current condition or performance of the nonvolatile solid-state device as a whole. As the nonvolatile solid-state device undergoes wear, performance of the hybrid HDD is maintained by using less reliable memory blocks in the nonvolatile solid-state device as a read cache, even when these memory blocks are considered too unreliable to store dirty data.

Hybrid-HDD with multiple caching modes

A data storage device that includes a magnetic storage device selects one or more power states of the magnetic storage device based on a time interval since a most recent time data has been read from or written to the magnetic storage device. The power state of the magnetic storage device can be changed from a higher power consumption state to a lower power consumption state when the time interval exceeds a predetermined value. The power consumption state may be changed from an active servo state to an intermediate power consumption state, a park state, and/or a standby state, depending on the time elapsed since the most recent time data has been read from or written to the magnetic storage device.

Power state change in disk drive based on disk access history

A hybrid drive and associated methods provide low-overhead storage of a hibernation file in the hybrid hard disk drive. During operation, the hybrid drive allocates a portion of solid-state memory in the drive that is large enough to accommodate a hibernation file associated with a host device of the hybrid drive. In addition to the erased memory blocks that are normally present during operation of the hybrid drive, the portion of solid-state memory allocated for accommodating the hibernation file may include over-provisioned memory blocks, blocks used to store a previous hibernation file that has been trimmed, and/or non-dirty blocks.

Low-overhead storage of a hibernation file in a hybrid disk drive

A hybrid drive and associated methods increase the rate at which data are transferred to a nonvolatile storage medium in the hybrid drive. By using a large nonvolatile solid state memory device as cache memory for a magnetic disk drive, a very large number of write commands can be cached and subsequently reordered and executed in an efficient manner. In addition, strategic selection and reordering of only a portion of the write commands stored in the nonvolatile solid state memory device increases efficiency of the reordering process.

Write reordering in a hybrid disk drive

In a cache policy for a hybrid drive having a magnetic storage device and a non-volatile solid-state device, the hybrid drive is configured to write the most recent version of data associated with a logical block address to the non-volatile solid-state device when the logical block address is associated with previously written data and is overlapped by a subsequent disk write operation. Advantageously, the most recent version of data associated with the overlapped logical block address is stored in cache in the non-volatile solid-state device, even when the subsequent disk write operation results in the overlapped logical block address being trimmed from cache or otherwise invalidated. Consequently, data associated with the overlapped logical block address can be accessed more quickly than data written to the magnetic storage device.

Hybrid-HDD policy for what host-R/W data goes into NAND

Data are refreshed in a nonvolatile solid-state device to significantly reduce the likelihood of data retention errors. Test data are written in a region of the nonvolatile solid-state device when user data are stored in the nonvolatile solid-state device, and are subsequently read to detect the possibility of data retention errors occurring when the associated user data are read. The test data may be a portion of the user data or a predetermined test pattern. To increase sensitivity to incipient charge leakage that may compromise the user data, the test data may be written using a modified write process and/or read with a modified read operation. The nonvolatile solid-state device may be employed as part of a solid-state drive or as the flash-memory portion of a hybrid hard disk drive.

Growing community of inventors

Cupertino, CA, United States of America

Eric R Dunn