Abstract: A disk drive is disclosed comprising a non-volatile write cache and a head actuated over a disk. A plurality of write commands are received from a host, wherein each write command comprises write data. A workload for a non-cache area of the disk is determined, and when the workload for the non-cache area of the disk is less than a threshold independent of a workload for the write cache, substantially all of the write data is stored in the non-cache area of the disk. When the workload for the non-cache area of the disk is greater than the threshold independent of the workload for the write cache, a first percentage of the write data is stored in the non-volatile write cache and a second percentage of the write data is stored in the non-cache area of the disk, wherein the first percentage is proportional to the workload for the non-cache area of the disk.

Abstract: A flash memory device is disclosed comprising a flash controller for accessing a first flash memory over a first channel and a second flash memory over a second channel. A multi-command descriptor block is received from a host, wherein the multi-command descriptor block comprises identifiers for identifying a plurality of access commands that the host is preparing to request. A first group of the access commands are selected to execute concurrently and a second group of the access commands are selected to execute concurrently. The first group of access commands are received from the host and executed concurrently by accessing at least the first and second flash memories concurrently. The second group of access commands are received from the host and executed concurrently by accessing at least the first and second flash memories concurrently.

Abstract: A non-volatile storage system is disclosed which provides a mapping table which includes a granularity which does not correspond to the page size of a non-volatile storage array. A reduced mapping table granularity enables more than one mapping entry to exist in a single page on the solid-state array. A write command which does not exceed a mapping table entry can invalidate only a portion of the written page, and can be combined with a second write command to write a new page of the solid-state array.

Abstract: A hybrid drive is disclosed comprising a head actuated over a disk, and a non-volatile semiconductor memory (NVSM). When a write command is received from a host that is mapped to the NVSM, the write command is serviced by writing data to the NVSM, and when a life remaining of the NVSM falls below a threshold, by also writing a copy of the data to the disk.

Abstract: The present invention is directed to systems and methods for improving access to non-volatile solid-state storage systems. Embodiments described herein provide a physical chunk number (PCN), or a physical page number (PPN), by which a controller can access the next available chunks (or pages) in a programming sequence optimized by concurrency. By incrementing the PCN, the controller can program consecutive chunks in the optimized programming sequence. In one embodiment, the programming sequence is determined at the time of initial configuration and the sequence seeks to synchronize data programming and data sending operations in subcomponents of the storage system to minimize contention and wait time. In one embodiment, the PCN includes an index portion to a superblock table with entries that reference specific blocks within the subcomponents in a sequence that mirrors the optimized programming sequence, and a local address portion that references a particular chunk to be programmed or read.

Abstract: A hybrid drive is disclosed comprising a head actuated over a disk, and a non-volatile semiconductor memory (NVSM). A first execution time needed to execute commands in a NVSM command queue is estimated, and a second execution time needed to execute commands in a disk command queue is estimated. An access command is inserted into a selected one of the NVSM command queue and the disk command queue in response to the first and second execution times, and one of the first and second execution times is updated in response to an estimated execution time of the access command.

Abstract: Embodiments of the invention are directed to enabling concurrent commands from command requesters in a non-volatile solid-state storage subsystem in a manner that reduces contentions among the commands. Embodiments group blocks of memory into multiple sets of superblocks and associate a command requester to each superblock set. In one embodiment, the superblock sets are dynamically associated with a requester. In one embodiment, the superblock sets are dynamically associated with requesters based in part on at least one of internal memory management needs and host command throughput. In one embodiment, an erase command is executed on a superblock within a set and a simultaneous write command is executed on a superblock within another set.

Abstract: A non-volatile semiconductor memory module is disclosed comprising a memory device and memory controller operably coupled to the memory device, wherein the memory controller is operable to receive a host command, split the host command into one or more chunks comprising a first chunk comprising at least one logical block address (LBA), and check the first chunk against an active chunk coherency list comprising one or more active chunks to determine whether the first chunk is an independent chunk, and ready to be submitted for access to the memory device, or a dependent chunk, and deferred access to the memory device until an associated dependency is cleared.

Abstract: Embodiments of the invention include systems and methods for reducing bus transfers for a storage device. In particular, these systems and methods reduce bus transfers by modifying an interface transfer protocol which designates the size of a multiple block read or write command is transmitted in a separate block transfer size command. Separate block transfer size commands can be omitted where the storage device maintains a record of a previously used block transfer size and reuses the size for subsequent multiple block read or write commands.

Abstract: Systems and methods are disclosed for recovering from a data access error encountered in data stripes implemented in a data redundancy scheme (e.g., RAID) in a solid state storage device. In one embodiment, the storage device holds parity data in a temporary, volatile memory such as a RAM and writes the parity data to the non-volatile memory when a full stripe's worth of new write data has been written to the non-volatile memory. In one embodiment, upon detecting that a data access error has occurred in a partially written stripe, the storage device initiates a write of the parity data for the partially written stripe to the non-volatile memory and executes a RAID recovery procedure using the newly written parity data to attempt to recover from the detected error. This approach allows for a recovery from the data access error without waiting for the full stripe to be written.

Abstract: In one embodiment of the invention, a flash-based/solid-state storage system with an implemented data redundancy scheme such as RAID is configured to hold parity data in a volatile memory such as RAM and write such parity data to the non-volatile flash media when a full stripe of data has been written to the media. Other embodiments in certain situations force an early write of the parity for a partial stripe that has not been fully written to the non-volatile media. Those situations may include a data access error on data in a partial stripe and a detected power loss event with a partial stripe present. Embodiments are directed to writing additional data with the parity data for the partial stripe and then later using the additional data in data recovery. This approach allows the controller to easily detect the presence of a partial stripe and handle such a stripe accordingly.

Abstract: A hybrid drive is disclosed comprising a head actuated over a disk, and a non-volatile semiconductor memory (NVSM). A write command is received from a host, the write command comprising data. First and second error correction code (ECC) symbols are generated over the data, wherein the second ECC symbols are different than the first ECC symbols. The data and first ECC symbols are written to the NVSM, and the second ECC symbols are written to the disk.

Abstract: A solid-state storage retention monitor determines whether user data in a solid-state device is in need of a scrubbing operation. One or more reference blocks may be programmed with a known data pattern, wherein the reference block(s) experiences substantially similar P/E cycling, storage temperature, storage time, and other conditions as the user blocks. The reference blocks may therefore effectively represent data retention properties of the user blocks and provide information regarding whether/when a data refreshing operation is needed.

Abstract: A hybrid drive is disclosed comprising a head actuated over a disk comprising a plurality of data tracks, where each data track comprises a plurality of data sectors. The hybrid drive further comprises a non-volatile semiconductor memory (NVSM) comprising a plurality of memory segments. When a write command is received from a host including write data, the write data is written to one of a disk cache and a NVSM cache, wherein the write data is eventually flushed to a non-cache area of the disk.

Abstract: Systems and methods for retaining data in non-volatile solid-state are disclosed in which refresh copy operations are performed on data stored in non-volatile solid-state memory. A controller can comprise a data retention module configured to issue copy commands within different periods of time, and to maintain usage data on a storage subsystem. A refresh copy operation helps ensure that data written to memory retain integrity by causing data to be programmed again onto the memory, which minimizes the risk of data error caused by electron leak in the non-volatile solid-state memory. One or more data structures may be used to determine memory blocks that require refresh copy operations. In one embodiment, a validity bit array is used to track blocks that contain valid data. In another embodiment, a least recently used list is used to track blocks that have been least recently written.

Abstract: A data storage system configured to manage unreliable memory units is disclosed. In one embodiment, the data storage system maintains an unreliable memory unit list designating memory units in a non-volatile memory array as reliable or unreliable. The unreliable memory unit list facilitates management of unreliable memory at a granularity level finer than the granularity of a block of memory. The data storage system can add entries to the unreliable memory unit list as unreliable memory units are discovered. Further, the data storage system can continue to perform memory access operations directed to reliable memory units in blocks containing other memory units determined to be unreliable. As a result, the operational life of the data storage system is extended.

Abstract: A solid state drive includes a plurality of flash memory devices, and a memory controller coupled to the plurality of flash memory devices. The memory controller is configured to logically associate blocks from the plurality of flash memory devices to form zip codes, the zip codes associated with corresponding erase counters. The solid state drive further includes a processor and a computer-readable memory having instructions stored thereon. The processor may perform a wear-leveling operation by determining that blocks in a first zip code have been erased and incrementing a first erase counter associated with the first zip code. It may then be determined that a second erase counter associated with a second zip code is low relative to at least one other erase counter, and based on this determination, data from blocks in the second zip code may be written to new blocks as part of a wear-leveling operation.

Abstract: The present invention is directed to systems and methods for improving access to non-volatile solid-state storage systems. Embodiments described herein provide a physical chunk number (PCN), or a physical page number (PPN), by which a controller can access the next available chunks (or pages) in a programming sequence optimized by concurrency. By incrementing the PCN, the controller can program consecutive chunks in the optimized programming sequence. In one embodiment, the programming sequence is determined at the time of initial configuration and the sequence seeks to synchronize data programming and data sending operations in subcomponents of the storage system to minimize contention and wait time. In one embodiment, the PCN includes an index portion to a superblock table with entries that reference specific blocks within the subcomponents in a sequence that mirrors the optimized programming sequence, and a local address portion that references a particular chunk to be programmed or read.

Abstract: Solid State Drives (SSD) can yield very high performance if it is designed properly. A SSD typically includes both a front end that interfaces with the host and a back end that interfaces with the flash media. Typically SSDs include flash media that is designed with a high degree of parallelism that can support a very high bandwidth on input/output (I/O). A SSD front end designed according to a traditional hard disk drive (HDD) model will not be able to take advantage of the high performance offered by the typical flash media. Embodiments of the invention provide improved management of multiple I/O threads that take advantage of the high performing and concurrent nature of the back end media, so the resulting storage system can achieve a very high performance.

Abstract: A data storage device is disclosed comprising a non-volatile memory including a plurality of memory segments. A write command is received comprising a logical block address (LBA) and user data. The LBA is mapped to a physical block address (PBA) for addressing one of the memory segments. First error code redundancy is generated in response to the LBA, and second error code redundancy in response to the PBA. User data and the first and second error code redundancy are written to the memory segment addressed by the PBA.