Avago Technologies General IP (singapore) Pte. Ltd.

Western Digital Technologies, Inc.

Avago Technologies International Sales Pte. Limited

Saugata Das Purkayastha

Kishore Kaniyar Sampathkumar

Luca Bert

Sumanesh Samanta

Anant Baderdinni

Parag Ranjan Maharana

Philip K Wong

Revanasiddaiah Prabhuswamy Mathada

Nitin Kumar Gupta

Mark Ish

Horia Cristian Simionescu

Amiya Banerjee

Jameer Mulani

Anindya Rai

Mohana Rao Goli

Chaitanya Kavirayani

Sourav Saha

Sampath Raja Murthy

Rishabh Dubey

Shreesha Prabhu

Aspects of the disclosure provide for management of a flash translation layer (FTL) for a non-volatile memory (NVM) in a Solid State Drive (SSD). The methods and apparatus provide a logical to physical (L2P) table where a first portion of the table is used for mapping frequently accessed hot data to a first subdrive in the NVM. Additionally, a second portion of the L2P table is provided for mapping cold data less frequently accessed than the hot data to a second subdrive, where logical blocks for storing the cold data in the second subdrive are larger than logical blocks storing the hot data in the first subdrive. Separation of the L2P table into hot and cold subdrives reduces the L2P table size that is needed in RAM for logical to physical memory mapping, while at the same time provides lower write amplification and latencies, especially for large capacity SSDs.

Methods and apparatus for variable size logical page management based on hot and cold data

A non-volatile storage apparatus includes a plurality of non-volatile memory cells and control circuitry. The control circuitry is configured to apply one or more soft erase pulses to the plurality of non-volatile memory cells to reduce threshold voltages of the plurality of non-volatile memory cells from initial levels corresponding to programmed data to intermediate levels below the initial levels and above an erased level. The control circuitry is configured to apply one or more soft programming pulse to increase threshold voltages of the plurality of non-volatile memory cells from the intermediate levels to final levels corresponding to the programmed data.

Soft erase and programming of nonvolatile memory

Apparatuses and techniques are provided for switching a solid-state device (SSD) between a traditional SSD mode and an open-channel SSD (OCSSD) mode. In one aspect, a set of commands are defined for communicating different types of data from a Flash Translation Layer (FTL) between the host and SSD. The commands can include different predetermined sets of bits which carry different types of FTL data in a standardized format. The commands can be transmitted using reserved data words in an interface specification which interfaces the host to the SSD. The commands can include commands to transfer a logical-to-physical address table, a validity bitmap and a wear table. A switch mode command can include a bit indicating whether the mode is to be switched from traditional SSD to OCSSD, or from OCSSD to traditional SSD.

Method to switch between traditional SSD and open-channel SSD without data loss

A non-volatile storage apparatus includes a set of non-volatile memory cells and one or more control circuits in communication with the set of non-volatile memory cells. The one or more control circuits are configured to receive a plurality of non-sequential memory access commands directed to the set of non-volatile memory cells, predict a predicted memory access command based on the plurality of non-sequential memory access commands, and access the set of non-volatile memory cells according to the predicted memory access command.

Non-volatile storage with adaptive command prediction

Applications that use non-volatile random access memory (NVRAM), such as those that apply file system journal writes and database log writes where write operations apply data sequentially over the NVRAM, map the available capacity of the NVRAM in a virtual address space without compromising performance. The NVRAM is segmented into regions with multiple such regions fitting within a volatile RAM element accessible to the application and the NVRAM. One or more regions are loaded in the volatile RAM and reflected in page tables that reference the regions. The page tables are managed on a host computer executing the application. One region space in the volatile RAM is unused and available for transferred information. Mechanisms are provided for dynamically transferring regions and interfacing with the host computer. As the application sequentially accesses information in the stored regions, older regions are removed and new regions loaded from NVRAM to the volatile RAM.

System, method and computer-readable medium for dynamically mapping a non-volatile memory store

Systems and methods presented herein provide for simulated NVDRAM operations. In a host system, a host memory is sectioned into pages. An HBA in the host system comprises a DRAM and an SSD. The DRAM and the SSD are also sectioned into pages and mapped to pages of the host memory. A host processor is operable to generate Input/Output (I/O) requests. An HBA driver is operable to process the I/O requests. The HBA driver is also operable to detect when the pages of the DRAM are accessed, to determine a rate of page reclamation based on the detection, and to reclaim pages of data in the DRAM by moving pages of data from the DRAM into the pages of the SSD based on the determined rate of page reclamation.

Caching systems and methods for page reclamation with simulated NVDRAM in host bus adapters

An apparatus comprising a memory and a controller. The memory may be configured to (i) implement a cache and (ii) store meta-data. The cache comprises one or more cache windows. Each of the one or more cache windows comprises a plurality of cache-lines configured to store information. Each of the plurality of cache-lines may be associated with meta-data indicating one or more of a dirty state and an invalid state. The controller may be connected to the memory and configured to detect an input/output (I/O) operation directed to a file system. The controller may perform a read-fill based on a hint value when there is a read miss in the cache. The hint value may be based on the application access pattern. The hint value may be passed to a caching layer with a corresponding I/O.

Enhanced SSD caching

Systems and methods presented herein provide for simulated NVDRAM operations. In a host system, a host memory is sectioned into pages. An HBA in the host system comprises a DRAM and an SSD for cache operations. The DRAM and the SSD are sectioned into pages and mapped to pages of the host memory. The SSD is further sectioned into regions comprising one or more pages of the SSD. AnHBA driver is operable to load a page of data from the SSD into a page of the DRAM when directed by a host processor, to determine that the page of the DRAM is occupied with other data, to determine a priority of the region of the page of other data occupying the page of the DRAM, and to flush the other data from the DRAM to the SSD based on the determined priority.

Caching systems and methods for execution within an NVDRAM environment

An apparatus comprising a memory and a controller. The memory may be configured to (i) implement a cache and (ii) store meta-data. The cache comprises one or more cache windows. Each of the one or more cache windows comprises a plurality of cache-lines configured to store information. Each of the cache-lines comprises a plurality of sub-cache lines. Each of the plurality of cache-lines and each of the plurality of sub-cache lines is associated with meta-data indicating one or more of a dirty state and an invalid state. The controller is connected to the memory and configured to (i) recognize sub-cache line boundaries and (ii) process the I/O requests in multiples of a size of said sub-cache lines to minimize cache-fills.

System and methods for caching a small size I/O to improve caching device endurance

An apparatus having an interface and a circuit is shown. The interface is configured to receive a request to access a memory. The request includes a hint. The circuit is configured to select a current one of a plurality of cache policies based on the hint. The current cache policy includes a number of current parameters ranging from some to all of a plurality of management parameters of a cache device. The circuit is further configured to cache data of the request based on the current cache policy.

Hint selection of a cache policy

Methods and structure for selective cache mirroring. One embodiment includes a control unit and a memory. The memory is able to store indexing information for a multi-device cache for a logical volume. The control unit is able to receive an Input/Output (I/O) request from a host directed to a Logical Block Address (LBA) of the logical volume, to consult the indexing information to identify a cache line for storing the I/O request, and to store the I/O request at the cache line on a first device of the cache. The control unit is further able to mirror the I/O request to another device of the cache if the I/O request is a write request, and to complete the I/O request without mirroring the I/O request to another device of the cache if the I/O request is a read request.

Selective mirroring in caches for logical volumes

A storage subsystem can achieve more efficient I/O processing by enabling users to specify and pass out of band I/O hints comprising an object to be hinted, a hint type, and caching strategies associated with a hint type. A hinted object may be either a virtual device or a file. In addition to priority cache, hint types may include never-cache, sticky-cache, and volatile-cache. Hints may be passed via command-line or graphical-user interfaces.

Systems and methods for I/O processing using out-of-band hinting to block driver or storage controller

An apparatus comprising a memory and a controller. The memory may be configured to (i) implement a cache and (ii) store meta-data. The cache may comprise one or more cache windows. Each of the one or more cache windows comprises a plurality of cache-lines configured to store information. The controller is connected to the memory and configured to (A) process normal read/write operations in a first mode and (B) process special read/write operations in a second mode by (i) tracking a write followed by read condition on each of said cache windows and (ii) discarding data on the cache-lines associated with the cache windows after completion of the write followed by a read condition on the cache-lines.

Growing community of inventors

Bengaluru, India

Saugata Das Purkayastha