摘要:
A hybrid solid-state disk (SSD) has multi-level-cell (MLC) or single-level-cell (SLC) flash memory, or both. SLC flash may be emulated by MLC that uses fewer cell states. A NVM controller converts logical block addresses (LBA) to physical block addresses (PBA). Most data is block-mapped and stored in MLC flash, but some critical or high-frequency data is page-mapped to reduce block-relocation copying. A hybrid mapping table has a first-level and a second level. Only the first level is used for block-mapped data, but both levels are used for page-mapped data. The first level contains a block-page bit that indicates if the data is block-mapped or page-mapped. A PBA field in the first-level table maps block-mapped data, while a virtual field points to the second-level table where the PBA and page number is stored for page-mapped data. Page-mapped data is identified by a frequency counter or sector count. SRAM space is reduced.
摘要:
Truncation reduces the available striped data capacity of all flash channels to the capacity of the smallest flash channel. A solid-state disk (SSD) has a smart storage switch salvages flash storage removed from the striped data capacity by truncation. Extra storage beyond the striped data capacity is accessed as scattered data that is not striped. The size of the striped data capacity is reduced over time as more bad blocks appear. A first-level striping map stores striped and scattered capacities of all flash channels and maps scattered and striped data. Each flash channel has a Non-Volatile Memory Device (NVMD) with a lower-level controller that converts logical block addresses (LBA) to physical block addresses (PBA) that access flash memory in the NVMD. Wear-leveling and bad block remapping are preformed by each NVMD. Source and shadow flash blocks are recycled by the NVMD. Two levels of smart storage switches enable three-level controllers.
摘要:
Truncation reduces the available striped data capacity of all flash channels to the capacity of the smallest flash channel. A solid-state disk (SSD) has a smart storage switch salvages flash storage removed from the striped data capacity by truncation. Extra storage beyond the striped data capacity is accessed as scattered data that is not striped. The size of the striped data capacity is reduced over time as more bad blocks appear. A first-level striping map stores striped and scattered capacities of all flash channels and maps scattered and striped data. Each flash channel has a Non-Volatile Memory Device (NVMD) with a lower-level controller that converts logical block addresses (LBA) to physical block addresses (PBA) that access flash memory in the NVMD. Wear-leveling and bad block remapping are preformed by each NVMD. Source and shadow flash blocks are recycled by the NVMD. Two levels of smart storage switches enable three-level controllers.
摘要:
A flash-card reader improves transmission efficiency by using bulk streaming of multiple pipes. A bulk data-out pipe carries host write data to the card reader and can operate in parallel with a bulk data-in pipe that carries host read data that was read from a flash card attached to the card reader. Status packets do not block data packets since the he status packets are buffered through a separate status pipe, and commands are buffered through a command pipe. Flash data from multiple flash cards are interleaved as separate endpoints that share the bulk data-in pipe. A data in/out streaming state machine controls streaming bulk data through the bulk data-in and data-out pipes, while a status streaming state machine controls streaming status packets through the status pipe. Transaction overhead is reduced using bulk streaming where packets for several commands are combined into the same bulk streams.
摘要:
A flash system has multiple channels of flash memory chips that can be accessed in parallel. Host data is assigned to one of the channels by a multi-channel controller processor and accumulated in a multi-channel page buffer. When a page boundary in the page buffer is reached, the page buffer is written to a target physical block if full, or combined with old data fragments in an Aggregating Flash Block (AFB) when the logical-sector addresses (LSA's) match. Thus small fragments are aggregated using the AFB, reducing erases and wear of flash blocks. The page buffer is copied to the AFB when a STOP command occurs. Each channel has one or more AFB's, which are tracked by an AFB tracking table.
摘要:
A flash module has raw-NAND flash memory chips accessed over a physical-block address (PBA) bus by a NVM controller. The NVM controller is on the flash module or on a system board for a solid-state disk (SSD). The NVM controller converts logical block addresses (LBA) to physical block addresses (PBA). Data striping and interleaving among multiple channels of the flash modules is controlled at a high level by a smart storage transaction manager, while further interleaving and remapping within a channel may be performed by the NVM controllers. A SDRAM buffer is used by a smart storage switch to cache host data before writing to flash memory. A Q-R pointer table stores quotients and remainders of division of the host address. The remainder points to a location of the host data in the SDRAM. A command queue stores Q, R for host commands.
摘要:
An electronic flash-memory card has additional pipes for commands and status messages so that data pipes are not clogged with commands and status messages, allowing for a higher data throughput. The command and status pipes are activated when a UAS/BOT detector detects that a host is using a USB-Attached-SCSI (UAS) mode rather than a Bulk-Only-Transfer (BOT) mode. The host can send additional commands and data without waiting for completion of a prior command when operating in UAS mode but not while operating in BOT mode. A command queue (CQ) in the device re-orders commands for accessing flash memory and merges data in a RAM buffer. Smaller 1 KB USB packets in the data pipes are merged into larger 8 KB payloads in the RAM buffer, allowing for more efficient flash access.
摘要:
An extended universal-serial bus (EUSB) bridge to a host computer can have peripheral component interconnect express (PCIE) protocol layers on one side of the bridge, and EUSB layers on the other side of the bridge, with a high-level bridging converter module connecting the upper layers. The PCIE physical, data-link, and transport layers may be eliminated by integrating the bridge with an I/O controller. PCIE requests and data payloads are directly sent to the bridge, rather than low-level PCIE physical signals. The PCIE data payloads are converted to EUSB data payloads by a high-level direct bridging converter module. Then the EUSB data payloads are passed down to an EUSB transaction layer, an EUSB data-link layer, and an EUSB physical layer which drives and senses physical electrical signals on both differential pairs of the EUSB bus.
摘要:
A flash memory solid-state-drive (SSD) has a smart storage switch that reduces write acceleration that occurs when more data is written to flash memory than is received from the host. Page mapping rather than block mapping reduces write acceleration. Host commands are loaded into a Logical-Block-Address (LBA) range FIFO. Entries are sub-divided and portions invalidated when a new command overlaps an older command in the FIFO. Host data is aligned to page boundaries with pre- and post-fetched data filling in to the boundaries. Repeated data patterns are detected and encoded by compressed meta-data codes that are stored in meta-pattern entries in a meta-pattern cache of a meta-pattern flash block. The sector data is not written to flash. The meta-pattern entries are located using a meta-data mapping table. Storing host CRC's for comparison to incoming host data can detect identical data writes that can be skipped, avoiding a write to flash.
摘要:
A flash memory solid-state-drive (SSD) has a smart storage switch that reduces write acceleration that occurs when more data is written to flash memory than is received from the host. Page mapping rather than block mapping reduces write acceleration. Host commands are loaded into a Logical-Block-Address (LBA) range FIFO. Entries are sub-divided and portions invalidated when a new command overlaps an older command in the FIFO. Host data is aligned to page boundaries with pre- and post-fetched data filling in to the boundaries. Repeated data patterns are detected and encoded by compressed meta-data codes that are stored in meta-pattern entries in a meta-pattern cache of a meta-pattern flash block. The sector data is not written to flash. The meta-pattern entries are located using a meta-data mapping table. Storing host CRC's for comparison to incoming host data can detect identical data writes that can be skipped, avoiding a write to flash.