Abstract: A storage system includes a NAND storage media and a nonvolatile storage media as a write buffer for the NAND storage media. The write buffer is partitioned, where the partitions are to buffer write data based on a classification of a received write request. Write requests are placed in the write buffer partition with other write requests of the same classification. The partitions have a size at least equal to the size of an erase unit of the NAND storage media. The write buffer flushes a partition once it has an amount of write data equal to the size of the erase unit.

Abstract: A storage system includes a NAND storage media and a nonvolatile storage media as a write buffer for the NAND storage media. The write buffer is partitioned, where the partitions are to buffer write data based on a classification of a received write request. Write requests are placed in the write buffer partition with other write requests of the same classification. The partitions have a size at least equal to the size of an erase unit of the NAND storage media. The write buffer flushes a partition once it has an amount of write data equal to the size of the erase unit.

Abstract: Examples include techniques for managing high priority (HP) and low priority (LP) write transaction requests by a storage device. An embodiment includes receiving, at a storage controller for a storage device, a write transaction request from a requestor to write data to one or more memory devices in the storage device. When the write transaction request is for a high priority (HP) write, coalescing the write data into a transaction buffer in a memory of the storage device, sending an acknowledgment for the write transaction request to the requestor, and writing the write data into the one or more memory devices. When the write transaction request is for a low priority (LP) write, writing the write data into the one or more memory devices, and then sending an acknowledgment for the write transaction request to the requestor.

Abstract: Read Quality of Service in a solid state drive is improved by allowing a host system communicatively coupled to the solid state drive to control garbage collection in the solid state drive. Through the use of controlled garbage collection, the host system can control when to start and stop garbage collection in the solid state drive and the number of NAND dies engaged in garbage-collection operations.

Abstract: Systems, apparatuses and methods may provide for technology that reads a lower page, one or more intermediate pages and a last page from a set of multi-level non-volatile memory (NVM) cells, wherein one or more of a lower read time associated with the lower page or a last read time associated with the last page is substantially similar to an intermediate read time associated with the one or more intermediate pages.

Abstract: Techniques and mechanisms for exchanging information between a solid state drive (SSD) and a write-in-place non-volatile memory via a host device. In an embodiment, access control information defines state of the SSD, where the access control information determines and/or is based on an access by the host device to other non-volatile memory of the SSD. The access control information includes address conversion information defining a correspondence of a logical address with a physical address for a location of the other non-volatile memory of the SSD. At least some of the access control information is stored by the SSD to the write-in-place non-volatile memory for later retrieval by the SSD. In another embodiment, the SSD signals that a commit operation is to be performed to flush any cached or buffered access control information into the write-in-place non-volatile memory.

Abstract: Examples include techniques for managing high priority (HP) and low priority (LP) write transaction requests by a storage device. An embodiment includes receiving, at a storage controller for a storage device, a write transaction request from a requestor to write data to one or more memory devices in the storage device. When the write transaction request is for a high priority (HP) write, coalescing the write data into a transaction buffer in a memory of the storage device, sending an acknowledgment for the write transaction request to the requestor, and writing the write data into the one or more memory devices. When the write transaction request is for a low priority (LP) write, writing the write data into the one or more memory devices, and then sending an acknowledgment for the write transaction request to the requestor.

Abstract: A machine readable storage medium containing program code that when processed by a processor causes a method to be performed a method is described. The method includes executing a wear leveling routine by servicing cold data from a first queue in a non volatile storage device to write the cold data. The method also includes executing a garbage collection routing by servicing valid data from a second queue in the non volatile storage device to write the valid data. The method also includes servicing host write data from a third queue in the non volatile storage device to write the host write data wherein the first queue remains fixed and is serviced at a constant rate so that a runtime size of the third queue is not substantially affected by the wear leveling routine.

Abstract: Technologies for combining logical-to-physical address updates include a data storage device. The data storage device includes a non-volatile memory to store data and a logical to physical (L2P) table indicative of logical addresses and associated physical addresses of the data. Additionally, the data storage device includes a volatile memory to store one or more bins. Each bin is indicative of a subset of entries in the L2P table. Further, the data storage device includes a controller to allocate a bin in the volatile memory, write a plurality of updates to a subset of entries of the L2P table to the bin, and write the bin to the L2P table in a single write operation. Other embodiments are also described and claimed.

Abstract: Systems, apparatuses and methods may provide for technology that reads a lower page, one or more intermediate pages and a last page from a set of multi-level non-volatile memory (NVM) cells, wherein one or more of a lower read time associated with the lower page or a last read time associated with the last page is substantially similar to an intermediate read time associated with the one or more intermediate pages.

Abstract: Techniques are disclosed for write suppression to improve endurance rating of non-volatile memories, such as QLC-NAND SSDs or other relatively slow, low endurance non-volatile memories. In an embodiment, an SSD is configured with a fast frontend non-volatile memory, a relatively slow lower endurance backend non-volatile memory, and a frontend manager that selectively transfers data from the fast memory to the slow memory based on transfer criteria. In operation, write data from the host is initially written to the fast memory by the frontend manager. The data is moved from the fast memory to the slow memory in bands. For each data band stored in the fast memory, the frontend manager tracks invalid data counts and data age. Only bands that still remain valid are transferred to the slow memory. After a given band has been fully transferred, it is erased and re-usable for other incoming writes by the frontend manager.

Abstract: Several systems and methods of chip select are described. In one such method, a device maintains two identifiers, (ID_a and ID_m). When the device receives a command, it examines the values of ID_a and ID_m relative to a third reference identifier (ID_s). If either ID_a or ID_m is equivalent to ID_s, the device executes the command, otherwise, the device ignores the command. By using two different identification methods, a system has options in choosing to activate devices, being able to selectively switch between selecting multiple devices and single devices in a quick manner. In another such method, a device may have a persistent area that stores identification information such as an ID_a. Thus, system functionality may remain independent from any defect/marginality associated with the physical or logical components required for initial ID_a assignment of all devices in the system.

Abstract: Systems, apparatuses and methods may provide for initiating an erase of a block of non-volatile memory in response to an erase command, wherein the block includes a plurality of sub-blocks. Additionally, a failure of the erase with respect to a first subset of the plurality of sub-blocks may be tracked on an individual sub-block basis, wherein the erase is successful with respect to a second subset of the plurality of sub-blocks. In one example, use of the second subset of the plurality of sub-blocks is permitted, whereas use of the first subset of the plurality of sub-blocks is prevented.

Abstract: Systems, apparatuses and methods may provide for technology that reads a lower page, one or more intermediate pages and a last page from a set of multi-level non-volatile memory (NVM) cells, wherein one or more of a lower read time associated with the lower page or a last read time associated with the last page is substantially similar to an intermediate read time associated with the one or more intermediate pages.

Abstract: A hardware acceleration block is configured to process via a dedicated pair of registers, a plurality of commands of each of a plurality of threads received from a compute complex. The hardware acceleration block receives successive commands that are separated by at least an amount of time, from a thread of the plurality of threads. The amount of time is adequate to process a command from the thread.

Abstract: Technologies for performing a data copy operation on a data storage device include storing a copy token in a power-fail-safe data structure that identifies the source address and destination address of the data copy operation, updating an address table to indicate that the source and destination addresses are involved in the data copy operation, and notifying a host requesting that data copy operation that the data copy operation has been completed prior to performing the data copy operation. The host may subsequently perform other tasks while the data storage device completes the data copy operation. During the data copy operation, data access requests to the source or destination addresses are blocked based on the address table. Additionally, should a power failure event occur, the power-fail-safe data structure is saved to non-volatile data storage so that the copy operation may be completed upon the next power-on event of the data storage device.

Abstract: A storage system includes a NAND storage media and a nonvolatile storage media as a write buffer for the NAND storage media. The write buffer is partitioned, where the partitions are to buffer write data based on a classification of a received write request. Write requests are placed in the write buffer partition with other write requests of the same classification. The partitions have a size at least equal to the size of an erase unit of the NAND storage media. The write buffer flushes a partition once it has an amount of write data equal to the size of the erase unit.

Abstract: Provided are an apparatus, system and method for offloading collision check operations in a memory storage device to a collision check unit. A collision check unit includes a collision table including logical addresses for pending Input/Output (I/O) requests. An I/O request is received to a target logical address addressing a block of data in the non-volatile memory. The logical address is sent to the collision check unit. Resources to transfer data with respect to the transfer buffer to data for the I/O request are allocated in parallel while the collision check unit is determining whether the collision table includes the target logical address. The collision check unit determines whether the collision table includes the target logical address and returns indication of whether the collision table includes the target logical address indicating that current data for the target logical address is already in the transfer buffer.

Abstract: Technologies for providing cross data storage device communication include a compute device to transmit, with a processor, a move request to a first data storage device. The first data storage device is to transmit, in response to the move request, a completion notification to the processor. Additionally, the compute device is to read, with the first data storage device, after transmitting the completion notification, a block of data from a first non-volatile memory of the first data storage device to a volatile memory of the compute device. The first data storage device is to transmit to the second data storage device a second move request to move the block of data. The second data storage device is to write the block of data from the volatile memory to a second non-volatile memory of the second data storage device.

Abstract: Several systems and methods of chip select are described. In one such method, a device maintains two identifiers, (ID_a and ID_m). When the device receives a command, it examines the values of ID_a and ID_m relative to a third reference identifier (ID_s). If either ID_a or ID_m is equivalent to ID_s, the device executes the command, otherwise, the device ignores the command. By using two different identification methods, a system has options in choosing to activate devices, being able to selectively switch between selecting multiple devices and single devices in a quick manner. In another such method, a device may have a persistent area that stores identification information such as an ID_a. Thus, system functionality may remain independent from any defect/marginality associated with the physical or logical components required for initial ID_a assignment of all devices in the system.