Alibaba Group Holding Limited

T-head (shanghai) Semiconductor Co., Ltd.

Alibaba (china) Co., Ltd.

Fei Xue

Hongzhong Zheng

Shuangchen Li

Dimin Niu

Wentao Wu

Xiang Gao

Yuhao Wang

Jifeng Wang

Jiajing Jin

Yuming Xu

Fei Sun

Lide Duan

Jiu Heng

Wei Han

Yuanwei Fang

Yen-Kuang Chen

Feng Zhu

Yangjie Zhou

A solid state drive (SSD) includes an NAND memory and an SSD controller. The SSD controller includes an interface coupled to a host machine, a nonvolatile memory controller coupled to the interface, and a processor coupled to the nonvolatile memory controller. The SSD controller is configured to: receive, via the interface, a write command from the host machine; process, by the nonvolatile memory controller, the write command; transmit, from the nonvolatile memory controller to the processor, a system message; process, by the processor according to Zoned Namespaces (ZNS) protocol, the system message; obtain, by the nonvolatile memory controller via the interface, host data for storage from the host machine; and write the host data to the NAND memory based on a result of processing the system message. Processing the system message by the processor and obtaining the host data by the nonvolatile memory controller are executed in parallel.

Programmable control system and method for solid state drive

An SSD includes an MRAM, an NAND memory, and an SSD controller. The SSD controller is configured to receive first data from a host machine, save the first data to an SSD data buffer, fetch the first data from the SSD data buffer and write the first data to the MRAM via the MRAM controller, determine, by the data allocation circuit based on a characteristic of the first data, whether to save the first data to the MRAM or the NAND memory, and in response to determining saving the first data to the NAND memory, read the first data from the MRAM, write the first data to the NAND memory, and erase the first data from the MRAM.

Hybrid solid state drive (SSD) architecture using MRAM and NAND for improved power loss protection and performance

This application describes systems and methods for facilitating memory access on flash drives. An example method may start with receiving a read command on a flash memory from a host specifying a logic block address (LBA). The flash memory may include a plurality of blocks grouped into a plurality of super blocks, and each of the plurality of blocks may include a plurality of pages. The method may further include determining a zone identification and an LBA offset based on the LBA; determining a flash physical address (FPA) corresponding to the LBA by accessing a mapping table stored in a random access memory (RAM) according to the zone identification and the LBA offset (e.g., the mapping table includes a plurality of FPAs arranged in a plurality of zones corresponding to the plurality of super blocks); and determining a page number and a block identification corresponding to the FPA.

Software-hardware combination method for internal mapping address query of zoned namespace

This application describes systems and methods for facilitating memory access in flash drives. An example method performed by a memory controller may include receiving, from a host, a write command comprising data to be written into a flash memory; splitting the data into a first portion and a second portion; storing the first portion into a static random-access memory (SRAM) in the memory controller; storing the second portion into a dynamic random-access memory (DRAM) communicatively coupled with the memory controller; initiating a configuration operation corresponding to the write command; fetching the first portion from the SRAM and the second portion from the DRAM in response to the flash translation layer indicating a ready status to store the data into the flash memory; combining the fetched first portion and the fetched second portion; and storing the combined first portion and the second portion into the flash memory.

Data cache scheme for high performance flash memories

Zero skipping sparsity techniques for reduced data movement between memory and accelerators and reduced computational workload of accelerators. The techniques include detection of zero and near-zero values on the memory. The non-zero values are transferred to the accelerator for computation. The zero and near-zero values are written back within the memory as zero values.

Zero skipping techniques for reducing data movement

Embodiments of the disclosure provide devices and methods for performing a top-k function. The device can include: a memory comprising a plurality of register files for storing the data elements, the plurality of register files comprising a parent register file and a first child register file associated with the parent register file, wherein the parent register file is associated with: first interface circuitry configured for reading a first parent data element from the parent register file and receiving a first child data element and a second child data element from the first child register file; and first comparison circuitry configured for updating the parent register file and the first child register file based on the first parent data element, the first child data element, and the second child data element according to a given principle.

Performing a top-k function using a binary heap tree

A system-in-package architecture in accordance with aspects includes a logic die and one or more memory dice coupled together in a three-dimensional slack. The logic die can include one or more global building blocks and a plurality of local building blocks. The number of local building blocks can be scalable. The local building blocks can include a plurality of engines and memory controllers. The memory controllers can be configured to directly couple one or more of the engines to the one or more memory dice. The number and type of local building blocks, and the number and types of engines and memory controllers can be scalable.

Scalable system-in-package architectures

A flash memory device includes a plurality of flash memory cell arrays, wherein: a flash memory cell array in the plurality of flash memory cell arrays comprises a plurality of layers of flash memory cell planes; and a flash memory cell plane includes a plurality of flash memory cells. The flash memory device further includes a logic circuitry coupled to the plurality of flash memory cell arrays, configured to perform operations using the plurality of flash memory cell arrays; and a sensing circuitry configured to access a corresponding flash memory cell plane among the plurality of flash memory cell planes.

Flash memory with improved bandwidth

A dual-model memory interface of a computing system is provided, configurable to present memory interfaces having differently-graded bandwidth capacity to different processors of the computing system. A mode switch controller of the memory interface controller, based on at least an arbitration rule written to a configuration register, switches the memory interface controller between a narrow-band mode and a wide-band mode. In each mode, the memory interface controller disables either a plurality of narrow-band memory interfaces of the memory interface controller according to a first bus standard, or a wide-band memory interface of the memory interface controller according to a second bus standard. The memory interface controller virtualizes a plurality of system memory units of the computing system as a virtual wide-band memory unit according to the second bus standard, or virtualizes a system memory unit of the computing system as a virtual narrow-band memory unit according to the first bus standard.

Dual-modal memory interface controller

The present disclosure relates to a method for controlling execution of a GEMM operation on an accelerator comprising multiple computation units, a first memory device, and a second memory device. The method comprises determining an execution manner of the GEMM operation, the execution manner comprising partition information of the GEMM operation and computation unit allocation information of the partitioned GEMM operation; generating one or more instructions to compute the partitioned GEMM operation on one or more allocated computation units; and issuing the one or more instructions to at least one of a first queue and a second queue, which enables at least one of a first local controller and a second local controller to execute the one or more instructions, wherein the first local controller and the second local controller are configured to control data movement between the computation units, the first memory device, and the second memory device.

Programmable and hierarchical control of execution of GEMM operation on accelerator

A method of scheduling instructions in a processing system comprising a processing unit and one or more co-processors comprises dispatching a plurality of instructions from a master processor to a co-processor of the one or more co-processors, wherein each instruction of the plurality of instructions comprises one or more additional fields, wherein at least one field comprises grouping information operable to consolidate the plurality of instructions for decomposition, and wherein at least one field comprises control information. The method also comprises decomposing the plurality of instructions into a plurality of fine-grained instructions, wherein the control information comprises rules associated with decomposing the plurality of instructions into the plurality of fine-grained instructions. Further, the method comprises scheduling the plurality of fine-grained instructions to execute on the co-processor, wherein the scheduling is performed in a non-sequential order.

Methods of breaking down coarse-grained tasks for fine-grained task re-scheduling

An apparatus comprises a bulk array of non-volatile memory cells on an integrated circuit die and an arithmetic logic unit on the die coupled to the bulk array. The arithmetic logic unit is operable to perform arithmetic logic operations on contents of the bulk array responsive to instructions received from outside of the die. The non-volatile memory cells may include NAND-type flash memory cells.

System and method for processing large datasets

A system comprises: a three-dimensional flash memory comprising a plurality of cells; and a controller coupled to the three-dimensional flash memory, configured to: select a block of cells in the three-dimensional flash memory; perform a matrix multiplication on the matrix stored in the block of cells, including performing a vector multiplication in a single sensing step; and output a matrix multiplication result. A matrix is stored in the block of cells.

Growing community of inventors

Sunnyvale, CA, United States of America

Fei Xue