Storage array, memory and electronic equipment

By grouping storage units in the storage array and setting switch tubes, the problem of row/column current interference in large-scale arrays is solved, the area efficiency and integration density of the storage array are improved, and the accuracy of data reading is ensured.

CN118969036APending Publication Date: 2024-11-15TRUTH MEMORY CORP
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411102430.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-12
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In large-scale storage arrays, due to the lack of a gate transistor in the memory cell, the current interference between rows and columns seriously affects the accuracy of circuit reading, and traditional storage arrays are difficult to improve integration density.

Method used

By setting at least one storage unit group in each column in the storage array, each storage unit group is provided with a first switch tube, and a second switch tube is provided on each row, the gate signal line and column selection are read through the corresponding row The signal line controls the switch tubes separately to reduce the number of first switch tubes in each column of memory cells and reduce row/column current interference.

Benefits of technology

Improves the area efficiency and integration density of the storage array, while reducing the row/column current interference in large-scale arrays, ensuring the accuracy of data reading.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118969036A_ABST
    Figure CN118969036A_ABST
Patent Text Reader

Abstract

The invention provides a storage array, a memory and electronic equipment, and relates to the technical field of storage, and the storage array comprises a plurality of storage units arranged in a matrix mode; each column of storage array comprises at least one storage unit group, the storage unit group comprises a plurality of storage units and a first switch tube, and each storage unit comprises a storage device; each row of storage array comprises a second switch tube; three ends of the first switch tube are respectively connected with the spin orbit moment layer of the memory device in the corresponding group, the corresponding column common source line and the corresponding column gating signal line; three ends of the second switch tube are respectively connected with the magnetic tunnel junction of the memory device in the corresponding row, the corresponding read bit line and the corresponding row read gating signal line; according to the invention, row / column current interference existing in a large-scale array can be reduced while the number of the switch tubes in the storage unit is reduced to improve area efficiency, and the accuracy of reading operation is ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of storage technology, and in particular to storage arrays, memories, and electronic devices. Background Art

[0002] Magnetic random access memory (MRAM) has attracted great attention in the academic and industrial fields due to its advantages such as non-volatility, radiation resistance, and low power consumption. Magnetic tunnel junction (MTJ) is the basic storage unit of magnetic random access memory MRAM.

[0003] The storage cell structure in the traditional storage array includes two gate transistors and a SOT-MRAM device. Among them, SOT-MRAM is a spin-orbit torque magnetic memory, and its full English name is Spin-orbit Torque MRAM. SOT-MRAM generally includes a magnetic tunnel junction and a spin-orbit torque layer (also known as SOT layer) adjacent to the free layer of the magnetic tunnel junction, so it is difficult to improve the integration density of the storage array.

[0004] At present, in order to improve storage density, some arrays use memory cells composed of a single gate transistor and a SOT-MRAM memory device. The transistor in the memory cell of this structure is located on the bit line of the cell spin orbit moment layer and is used to control the write current. This structure improves the area efficiency of the array by reducing the MTJ top electrode transistor. However, as the array size continues to increase, due to the lack of a gate transistor in the memory cell, the current interference between the rows / columns of the array will seriously affect the accuracy of circuit reading. Summary of the invention

[0005] The present application provides a storage array, a memory and an electronic device, which are used to provide a technical solution for improving array area efficiency while reducing inter-row current interference in a large-scale array.

[0006] In a first aspect, the present application provides a storage array, comprising: a plurality of storage units arranged in a matrix;

[0007] Each column of the storage array includes at least one storage cell group, the storage cell group includes a plurality of storage cells and a first switch tube, and each storage cell includes a storage device;

[0008] Each row of the storage array includes a second switch tube;

[0009] The three ends of the first switch tube are respectively connected to the spin orbit moment layer of the storage device in the corresponding group, the corresponding column common source line and the corresponding column selection signal line;

[0010] The three ends of the second switch tube are respectively connected to the magnetic tunnel junction of the storage device in the corresponding row, the corresponding read bit line and the corresponding row read selection signal line;

[0011] When reading data, the second switch tube is selected by the signal in the corresponding row read selection signal line, and the first switch tube is selected by the signal in the corresponding column selection signal line. The storage device is used to read out the data in the corresponding storage device through the corresponding read bit line when the corresponding first switch tube and the corresponding second switch tube are turned on at the same time.

[0012] In an optional implementation, each of the storage units further includes a third switch tube, three ends of which are respectively connected to a corresponding write bit line, a first end of the corresponding storage device, and a corresponding row write word line;

[0013] When data is written, the third switch tube is selected by the signal in the corresponding row write word line, and the first switch tube is selected by the signal in the corresponding column selection signal line. The storage device is used to write data to the corresponding storage device through the corresponding write bit line and the corresponding column common source line when the corresponding third switch tube and the corresponding first switch tube are turned on.

[0014] In an optional implementation, in the data reading phase, the interference voltage accumulated by the plurality of memory cells in the memory cell group on the read bit line is less than or equal to a preset interference voltage.

[0015] In an optional implementation, the preset interference voltage is a maximum interference voltage that the read bit line can withstand.

[0016] In an optional embodiment, the storage device is a spin-orbit torque magnetic memory; the spin-orbit torque magnetic memory includes at least a magnetic tunnel junction and a spin-orbit moment layer; the magnetic tunnel junction includes at least a free layer, a barrier layer and a reference layer stacked in sequence, and the free layer is adjacent to the spin-orbit moment layer.

[0017] In a second aspect, the present application provides a memory, which includes a row address decoding module, a column address decoding module, a reading module, a column common source line, a column selection signal line, a read bit line, a row read selection signal line and the storage array described in the first aspect.

[0018] The column address decoding module is connected to the control end of the first switch tube through the corresponding column selection signal line, and the row address decoding module is connected to the control end of the second switch tube through the corresponding row read selection signal line; the reading module is connected to the first end of the second switch tube through the corresponding read bit line, and the second end of the second switch tube is connected to the magnetic tunnel junction of the storage device in the corresponding row.

[0019] When reading data, the row address decoding module selects the second switch tube through the signal in the corresponding row read selection signal line, and the column address decoding module selects the first switch tube through the signal in the corresponding column selection signal line. The reading module is used to read out the data in the memory through the corresponding read bit line when the first switch tube and the second switch tube are turned on at the same time.

[0020] In an optional embodiment, the memory also includes a write module, a write bit line, and a row write word line; each of the storage units also includes a third switch tube, the three ends of the third switch tube are respectively connected to the corresponding write bit line, the first end corresponding to the storage device and the corresponding row write word line.

[0021] The write module is connected to the storage device through the corresponding write bit line and the third transistor; the row address decoding module is connected to the control end of the third transistor through the corresponding row write word line.

[0022] When data is written, the row address decoding module selects the third switch tube through the signal in the corresponding row write word line, and the column address decoding module selects the first switch tube through the signal in the corresponding column selection signal line. The write module is used to write data to the storage device through the corresponding write bit line and the corresponding column common source line when the corresponding third switch tube and the corresponding first switch tube are turned on.

[0023] In an optional implementation, the column address decoder is used to generate a column enable signal according to an address signal, and the column enable signal is a signal in the corresponding column enable signal line.

[0024] The row address decoder is used to generate a row write signal according to the address signal, and the row write signal is a signal in the corresponding row write word line.

[0025] The row address decoder is further used to generate a row read enable signal according to the address signal, and the row read enable signal is a signal in the corresponding row read enable signal line.

[0026] In an optional implementation, the interference voltage accumulated by the plurality of memory cells in the memory cell group on the corresponding read bit line is less than or equal to a preset interference voltage.

[0027] In a third aspect, the present application provides an electronic device, comprising the memory described in the second aspect.

[0028] The technical solution provided by the present application is that when reading data, the second switch tube that needs to be turned on is selected by the signal in the corresponding row read selection signal line, and the first switch tube that needs to be turned on is selected by the signal in the corresponding column selection signal line, which is equivalent to turning on the first switch tube and the second switch tube in the row and column where the storage device is located in each reading operation. In the present application, the multiple storage cells in each column of the storage array are divided into at least one storage cell group, and each storage cell group is correspondingly provided with a first switch tube. Based on this, the number of first switch tubes in each column of storage cells can be reduced, thereby reducing the row / column current interference in the array.

[0029] In addition, the present application divides the multiple storage cells in each column of the storage array into at least one storage cell group, and each storage cell group is correspondingly provided with a first switch tube, which can reduce the number of switch tubes in the traditional storage array and improve the area efficiency of the storage array.

[0030] Therefore, the present application can reduce the number of switch tubes in the storage unit to improve area efficiency, while reducing the row / column current interference existing in large-scale arrays, thereby ensuring the accuracy of data reading. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0032] Figure 1 A schematic diagram of the structure of a storage array provided in an embodiment of the present application;

[0033] Figure 2 A schematic diagram of a storage array during data reading provided by an embodiment of the present application;

[0034] Figure 3 A schematic diagram of a storage array provided in an embodiment of the present application when data is written;

[0035] Figure 4 A schematic diagram of the structure of a memory provided in an embodiment of the present application.

[0036] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0037] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used herein in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. It is understood that the terms "first", "second", etc. used in the present application can be used to describe various information or data in this article, but these elements are not limited by these terms. These terms are only used to distinguish the first information from another information. For example, without departing from the scope of the present application, the first action information can be referred to as the second action information, and similarly, the second action information can be referred to as the first action information. Both the first action information and the second action information are action information, but they are not the same action information.

[0039] Magnetic random access memory (MRAM) has attracted great attention in the academic and industrial fields due to its advantages such as non-volatility, radiation resistance, and low power consumption. Magnetic tunnel junction is the basic storage unit of magnetic random access memory (MRAM).

[0040] The memory cell structure in the traditional memory array includes two gate transistors and a SOT-MRAM device. SOT-MRAM generally includes a magnetic tunnel junction and a spin-orbit moment layer (also known as a SOT layer) adjacent to the free layer of the magnetic tunnel junction, so it is difficult to improve the integration density of the memory array.

[0041] At present, in order to improve storage density, some arrays use memory cells composed of a single gate transistor and a SOT-MRAM memory device. The transistor in the memory cell of this structure is located on the bit line of the cell spin orbit moment layer and is used to control the write current. This structure improves the area efficiency of the array by reducing the MTJ top electrode transistor. However, as the array size continues to increase, due to the lack of a gate transistor in the memory cell, the current interference between the rows / columns of the array will seriously affect the accuracy of circuit reading.

[0042] In response to the above problems, the present application proposes a storage array, a memory and an electronic device to solve the technical problem that when the array uses a storage unit composed of a single selection transistor and a SOT-MRAM storage device, as the array scale continues to increase, the current interference between rows / columns of the array will seriously affect the accuracy of circuit reading due to the lack of a selection transistor in the storage unit.

[0043] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0044] First, refer to Figure 1 The embodiment of the present application provides a memory array, comprising: a plurality of memory cells arranged in a matrix; each column of the memory cells comprises at least one memory cell group 10, the memory cell group 10 comprises a plurality of memory cells 101 and a first switch tube S1, each memory cell 101 comprises a memory device. Each row of the memory array comprises a second switch tube S2.

[0045] That is to say, in the embodiment of the present application, each storage unit group 10 is correspondingly provided with a first switch tube S1, and each row of the storage array is correspondingly provided with a second switch tube S2.

[0046] The three terminals of the first switch tube S1 are respectively connected to the spin-track moment layer of the storage device in the corresponding group, the corresponding column common source line WSL and the corresponding column selection signal line SL.

[0047] More specifically, a first end of the first switch tube S1 is connected to one end of the spin-track moment layer of the plurality of storage devices in the corresponding group, a second end is connected to the corresponding column selection signal line SL, and a control end is connected to the corresponding column common source line WSL.

[0048] The three terminals of the second switch tube S2 are respectively connected to the magnetic tunnel junction of the memory device in the corresponding row, the corresponding read bit line RBL and the corresponding row read selection signal line WRBL.

[0049] More specifically, a first end of the second switch tube S2 is connected to the magnetic tunnel junction of the memory device in the corresponding row, a second end is connected to the corresponding read bit line RBL, and a control end is connected to the corresponding row read selection signal line WRBL.

[0050] Optionally, the number of memory cell groups 10 included in the memory array in the column direction in this embodiment can be set according to the number of memory cells in each column and the number of memory cells that can be set in the group memory cell. For example, when the column memory cell includes 60 memory cells and the group memory cell can be set to 6 memory cells, the column memory cell includes 10 groups of memory cells.

[0051] In the embodiment of the present application, the first switch tube S1 can be any suitable switch tube, and this embodiment does not specifically limit this. For example, the first switch tube S1 can be a bipolar junction transistor (BJT), a metal oxide semiconductor field effect transistor (MOSFET), an insulated gate bipolar transistor (IGBT), a field effect transistor (JFET), and the like.

[0052] Similarly, in the embodiment of the present application, the second switch tube S2 can be any suitable switch tube, and this embodiment does not specifically limit this. For example, the second switch tube S2 can be a bipolar transistor (BJT), a metal oxide semiconductor field effect transistor (MOSFET), an insulated gate bipolar transistor (IGBT), a field effect transistor (JFET), and the like.

[0053] Based on the above description, the embodiment of the present application sets a first switch tube S1 for each storage cell group 10 and sets a second switch tube S2 for each row of storage cells, which can improve the area efficiency of the storage array to a certain extent, thereby improving the integration density of the storage array.

[0054] Based on the above structure, refer to Figure 2 In the data reading stage, the second switch tube S2 is used to be selected according to the signal in the row read selection signal line WRBL, the first switch tube S1 is used to be selected according to the signal in the column selection signal line WSL, and the memory device is used to read out the data in the memory device through the corresponding read bit line RBL when the corresponding first switch tube S1 and the corresponding second switch tube S2 are turned on.

[0055] Based on the above description, in the reading operation, the embodiment of the present application only needs to turn on the second switch tube of the corresponding row and the first switch tube of the corresponding column. In this way, it is equivalent to turning on only at least one first switch tube in a storage unit group and a second transistor in the corresponding row each time data is read, which reduces the row / column current interference existing in large-scale arrays. Therefore, the embodiment of the present application can reduce the number of switches in the storage unit to improve the area efficiency while ensuring the accuracy of the reading operation.

[0056] The above provides the functions and structures of the storage array when reading data. When writing data to the storage array, refer to Figure 1Each of the storage units further comprises a third switch tube S3. The three terminals of the third switch tube S3 are respectively connected to a corresponding write bit line WBL, a spin orbit moment layer of the corresponding storage device and a corresponding row write word line WWL.

[0057] Specifically, the control end of the third switch tube S3 is connected to the corresponding row write word line WWL, the first end is connected to the corresponding write bit line WBL, the second end is connected to the other end of the spin-orbit moment layer corresponding to the storage device, and the other end of the spin-orbit moment layer corresponding to the storage device is connected to one end of the first transistor S1.

[0058] In the embodiment of the present application, the third switch tube S3 can be any suitable switch tube, and this embodiment does not specifically limit this. For example, the third switch tube S3 can be a bipolar junction transistor (BJT), a metal oxide semiconductor field effect transistor (MOSFET), an insulated gate bipolar transistor (IGBT), a field effect transistor (JFET), and the like.

[0059] Reference Figure 3 In the data writing stage, the third switch tube S3 is used to be selected according to the signal in the row write word line WWL, the first switch tube S1 is used to be selected according to the signal in the column selection signal line WSL, and the second switch tube S2 is used to be turned off according to the signal in the row read selection signal line WRBL. The memory device is used to write data to the corresponding memory device through the corresponding write bit line WBL and the corresponding column common source line SL when the corresponding first switch tube S1 and the corresponding third switch tube S3 are turned on.

[0060] Based on the above technical solution, when writing data, the embodiment of the present application only needs to turn on the third switch tube S3 of the corresponding row according to the signal in the row write word line WWL, and turn on the first switch tube S1 of the corresponding column according to the signal in the column selection signal line WSL, so as to realize data writing to the corresponding storage device.

[0061] During the data writing process, the write bit line WBL and the column common source line SL corresponding to the storage unit to be written need to be charged / discharged. It should be understood that since the embodiment of the present application groups each column of storage cells, the connection between the corresponding write bit line WBL and the corresponding column common source line SL to other storage devices is blocked by the unopened third switch tube in other rows, so the overall parasitic capacitance is reduced. Therefore, the parasitic capacitance on the write bit line WBL and the column common source line SL during the data writing process is small, thereby improving the efficiency of data writing.

[0062] The above describes the structure of the storage array and the operation process of the data writing stage and the data reading stage. This embodiment reduces one switch tube in each storage unit and configures a first switch tube for each storage unit group to improve the integration density of the array and ensure the accuracy of the reading operation. It should be understood that the number of storage units in each group of storage units is related to the integration density of the array and the accuracy of the reading operation. The following describes how to determine the number of storage units in each group.

[0063] The starting point of the embodiment of the present application is to balance the integration density of the memory array and the problem of inter-row current interference in large-scale arrays. Therefore, when determining the number of memory cells in each memory cell group, the embodiment of the present application should focus on the interference problem of the memory cells in the group.

[0064] In one example, an embodiment of the present application can determine the number of memory cells in a memory cell group by determining that, during a data reading phase, an interference voltage accumulated by multiple memory cells in the memory cell group on the corresponding read bit line is less than or equal to a preset interference voltage.

[0065] Exemplarily, when reading, the interference voltage is preset to 5 mV, and the voltage accumulated on the corresponding read bit line of each memory cell in the memory cell group is 0.5 mV-1 mV, then the number of memory cells in each group should not exceed 5 to 10.

[0066] Optionally, the preset interference voltage is a maximum interference voltage that the corresponding read bit line can withstand.

[0067] It should be noted that the above-mentioned maximum interference voltage that the corresponding read bit line can withstand may be the lower limit value of the interference voltage on the read bit line that affects the accuracy of the memory array reading.

[0068] Based on the above description, this embodiment can increase the number of storage cells in each group of storage cells as much as possible while ensuring the accuracy of storage array reading, so as to reduce the number of first switch tubes as much as possible, thereby improving the array integration density.

[0069] The above content is described based on the structure and function of the memory array, and the following content describes the structure of the memory device.

[0070] Before that, we first introduce MRAM, MTJ and SOT-MRAM.

[0071] MRAM is a non-volatile memory technology that uses the properties of magnetic materials to store data. MRAM combines the advantages of traditional magnetic storage and semiconductor storage, and has the characteristics of high speed, high density, low power consumption and high durability.

[0072] MTJ can present two states, that is, the magnetization directions of the two magnetic layers are parallel or antiparallel to each other, making the MTJ appear in a low resistance state or a high resistance state. This effect is called tunneling magnetoresistance effect. The basic principle of MRAM storage is to use the tunneling magnetoresistance effect, that is, high and low resistance states represent data "0" and "1" respectively. At present, there are two main technical routes for MRAM, including spin transfer torque magnetic random access memory devices (STT-MRAM) and spin orbit torque magnetic random access memory devices (SOT-MRAM).

[0073] SOT-MRAM devices are a new type of non-volatile memory technology that combines spin electronics and magnetic storage technology, and have the advantages of high speed, low power consumption and high durability.

[0074] In the embodiment of the present application, the SOT-MRAM device includes at least a magnetic tunnel junction and a spin-orbit moment layer.

[0075] The magnetic tunnel junction at least includes a free layer, a barrier layer and a reference layer stacked in sequence, and the free layer is adjacent to the spin-orbit moment layer.

[0076] In some examples, the magnetic tunnel junction of the SOT-MRAM device is the third terminal of the memory device, that is, in the embodiment of the present application, the magnetic tunnel junction of the SOT-MRAM device is connected to the corresponding read bit line through the second switch tube.

[0077] Based on this, the storage cells in the corresponding row can be selected and operated. For example, the storage cells in the corresponding row can be activated through the corresponding read bit line, so that the magnetic tunnel junctions in the corresponding row are connected to the read circuit to realize data reading.

[0078] In some other examples, the first end of the spin-orbit moment layer is the first end of the memory device, and the second end of the spin-orbit moment layer is the second end of the memory device. In other words, the first end of the spin-orbit moment layer of the SOT-MRAM device is connected to the corresponding first switch tube, and the second end is connected to the corresponding third switch tube.

[0079] Based on the above structure, when reading data, the first switch tube S1 of the corresponding row is activated through the corresponding column selection signal line WSL, so that the SOT-MRAM device of the corresponding column is connected to the external reading circuit. The second switch tube S2 of the corresponding row is activated through the corresponding row read selection signal line WRBL, so that the SOT-MRAM device of the corresponding row is connected to the external reading circuit, and the data in the corresponding storage device is read out through the corresponding read bit line RBL.

[0080] When writing data, the third switch tube S3 of the corresponding row is activated through the row write word line WWL, so that the storage unit of the corresponding row is connected to the write circuit. The first switch tube S1 of the corresponding column is activated through the column selection signal line WSL, so that the storage unit of the corresponding column is connected to the write circuit. Data is written to the corresponding storage device through the corresponding write bit line WBL and the corresponding column common source line SL.

[0081] Based on the above description, this example realizes precise control of data reading and data writing of corresponding storage cells through the combination of row write word line WWL and column strobe signal line WSL and the combination of row read strobe signal line WRBL and column strobe signal line WSL. In addition, the combination of row write word line and column strobe signal line and the combination of row read strobe signal line and column strobe signal line in this embodiment reduce unnecessary current paths, reduce current density, and improve device reliability and life. It has broad application prospects in the fields of cache, embedded storage and non-volatile storage.

[0082] The following is a more detailed introduction to the storage array provided in the embodiment of the present application using a specific embodiment.

[0083] The memory cells in the memory array in the embodiment of the present application adopt a 1T-1R design, that is, each memory cell includes: a third switch tube and a SOT-MRAM device, the third switch tube is located between the spin orbit moment layer and the write bit line WBL of the SOT-MRAM device, and is controlled by the WWL signal of each row. This structure reduces a gate transistor in the traditional SOT-MRAM memory device structure and improves the area efficiency of the array.

[0084] However, in order to reduce the row-to-row current interference in large-scale arrays while reducing the number of gate transistors, the embodiment of the present application groups the memory cells of multiple rows in the same column, and several memory cells in each group share a first switch tube, one end of which is connected to the column common source line SL, and the first switch tube is controlled by the column selection signal WSL. In addition, after the magnetic tunnel junctions of the SOT-MRAM devices in the same row are connected to each other, they are connected to one end of the second switch tube for row selection, and the other end of the second switch tube is connected to the read bit line RBL, and the second switch tube is controlled by the row read selection signal WRBL.

[0085] During the read operation, only the second switch tube of the corresponding row and the first switch tube of the corresponding column need to be turned on, and the third switch tube of each row needs to be turned off; during the write operation, the third switch tube of the corresponding row and the first switch tube in the group of the corresponding column need to be turned on, and the second switch tube of each row needs to be turned off. In this way, it is equivalent to turning on only a number of switches in a group during each read operation, reducing the row / column current interference in large-scale arrays, reducing the number of unit transistors to improve area efficiency, and ensuring the accuracy of the read operation.

[0086] The solution in the embodiment of the present application is applicable to large-scale integrated storage arrays, and because the SOT-MRAM device has the characteristics of non-volatile storage, it is applicable to low-power storage scenarios that require data retention in the event of power failure, such as end-side applications. Since the array design greatly improves the area efficiency, it is also particularly suitable for storage scenarios that require ultra-low area.

[0087] Second, refer to Figure 4 The embodiment of the present application also provides a memory, which includes a row address decoding module, a column address decoding module, a reading module, a column common source line SL, a column selection signal line WSL, a read bit line RBL, a row read selection signal line WRBL and the storage array described in the first aspect.

[0088] The column address decoding module is connected to the control end of the first switch tube S1 through the corresponding column selection signal line WSL, and the row address decoding module is connected to the control end of the second switch tube S2 through the corresponding row read selection signal line WRBL; the reading module is connected to the first end of the second switch tube through the corresponding read bit line RBL, and the second end of the second switch tube is connected to the magnetic tunnel junction of the storage device in the corresponding row.

[0089] When reading data, the row address decoding module enables the second switch tube S2 through the signal in the corresponding row read selection signal line WRBL, and the column address decoding module enables the first switch tube S1 through the signal in the corresponding column selection signal line WSL. The reading module is used to read out the data in the memory through the corresponding read bit line RBL when the first switch tube S1 and the second switch tube S2 are turned on at the same time.

[0090] That is to say, when the memory is in the reading stage, the reading module, the column address decoding module and the row address decoding module are activated, and the writing module does not work. The column address decoding module controls the signal in the WSL signal line to select the first switch tube in the group of the corresponding column according to the external input address signal; the row address decoding module is in the reading address decoding mode according to the external input mode signal, the signal in the WWL signal line of each row is pulled low, and the signal in the WRBL signal line of the corresponding row is pulled high, so that the second transistor of the corresponding row is turned on, thereby opening the reading path of the reading row, and the reading module reads the corresponding storage unit.

[0091] Based on the above description, during the reading operation, the embodiment of the present application only needs the column address decoding module to control the signal in the WSL signal line to select the first switch tube in the group corresponding to the column according to the external input address signal, and the row address decoding module to control the second transistor of the corresponding row to turn on according to the external input mode signal. In this way, it is equivalent to turning on only the first switch tube in a storage unit group and the second transistor of the corresponding row during each reading operation, which reduces the row / column current interference existing in large-scale arrays. Therefore, the embodiment of the present application can reduce the number of switches in the storage unit to improve the area efficiency while ensuring the accuracy of the reading operation.

[0092] It is worth noting that, in the data reading phase, the interference voltage accumulated by the plurality of memory cells in the memory cell group on the corresponding read bit line is less than or equal to a preset interference voltage.

[0093] Optionally, the preset interference voltage is a maximum interference voltage that the corresponding read bit line can withstand.

[0094] Furthermore, the maximum interference voltage that the corresponding read bit line can withstand may be a lower limit value of the interference voltage on the read bit line that affects the accuracy of reading the memory array.

[0095] Based on the above description, the embodiment of the present application can increase the number of storage cells in each group of storage cells as much as possible while ensuring the accuracy of storage array data reading, so as to minimize the number of first switch tubes and thereby improve the array integration density.

[0096] In an alternative example, refer to Figure 4 The memory also includes a write module, a write bit line WWL, and a row write word line WBL. Each of the storage units also includes a third switch tube S3, the three ends of which are respectively connected to the corresponding write bit line WWL, the first end of the corresponding storage device, and the corresponding row write word line WBL.

[0097] The write module is connected to the memory device through the corresponding write bit line WBL and the third transistor; the row address decoding module is connected to the control end of the third transistor S3 through the corresponding row write word line WWL;

[0098] When data is written, the row address decoding module selects the third switch tube S3 through the signal in the corresponding row write word line WWL, and the column address decoding module selects the first switch tube S1 through the signal in the corresponding column selection signal line WSL. The write module is used to write data to the storage device through the corresponding write bit line WWL and the corresponding column common source line SL when the corresponding third switch tube S3 and the corresponding first switch tube S1 are turned on.

[0099] That is to say, when the memory is in the writing stage, the writing module, the column address decoding module and the row address decoding module are activated, and the reading module does not work. The column address decoding module controls the signal in the WSL signal line to select the first switch tube S1 in the group of the corresponding column according to the external input address signal; the row address decoding module also controls the third switch tube S3 of the corresponding row through the signal in the WWL signal line according to the external input address signal, so that the writing current path is opened; at the same time, the writing module generates a writing voltage or current and applies it to WBL and SL, thereby completing the writing operation.

[0100] Based on the above technical solution, in the data writing stage, the embodiment of the present application only needs the row address decoding module to control the third switch tube of the corresponding row through the signal in the WWL signal line according to the external input address signal, and the column address decoding module to control the signal in the WSL signal line to select the first switch tube in the group of the corresponding column according to the external input address signal, so as to realize data writing to the target storage device.

[0101] Optionally, the column address decoder in the embodiment of the present application is used to generate a column enable signal according to an address signal, and the column enable signal is a signal in the corresponding column enable signal line.

[0102] The row address decoder is used to generate a row write signal according to the address signal, and the row write signal is a signal in the corresponding row write word line. The row address decoder is also used to generate a row read select signal according to the address signal, and the row read select signal is a signal in the corresponding row read select signal line. The above-mentioned address signal can be an externally input address signal, and the address signal can be determined according to the address of the storage device where data is written or read, and the embodiment of the present application does not make any special limitation on this.

[0103] In a third aspect, an embodiment of the present application provides an electronic device, which includes the memory in the second aspect.

[0104] Since the memory in the second aspect has significant advantages in performance, power consumption and reliability, the electronic device may include an electronic device suitable for application scenarios such as high-performance computing, embedded systems and industrial control.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A storage array, characterized in that: include: A plurality of storage units arranged in a matrix; Each column of the storage array includes at least one storage cell group, the storage cell group includes a plurality of storage cells and a first switch tube, and each storage cell includes a storage device; Each row of the storage array includes a second switch tube; The three ends of the first switch tube are respectively connected to the spin orbit moment layer of the storage device in the corresponding group, the corresponding column common source line and the corresponding column selection signal line; The three ends of the second switch tube are respectively connected to the magnetic tunnel junction of the storage device in the corresponding row, the corresponding read bit line and the corresponding row read selection signal line; When reading data, the second switch tube is selected by the signal in the corresponding row read selection signal line, and the first switch tube is selected by the signal in the corresponding column selection signal line. The storage device is used to read out the data in the corresponding storage device through the corresponding read bit line when the corresponding first switch tube and the corresponding second switch tube are turned on at the same time.

2. The storage array according to claim 1, characterized in that: Each of the storage units further comprises a third switch tube, wherein three ends of the third switch tube are respectively connected to a corresponding write bit line, a spin orbit moment layer of the corresponding storage device and a corresponding row write word line; When data is written, the third switch tube is selected by the signal in the corresponding row write word line, and the first switch tube is selected by the signal in the corresponding column selection signal line. The storage device is used to write data to the corresponding storage device through the corresponding write bit line and the corresponding column common source line when the corresponding third switch tube and the corresponding first switch tube are turned on.

3. The storage array according to claim 1, characterized in that: In the data reading phase, the interference voltage accumulated by the plurality of memory cells in the memory cell group on the corresponding read bit line is less than or equal to a preset interference voltage.

4. The storage array according to claim 3, characterized in that: The preset interference voltage is a maximum interference voltage that the corresponding read bit line can withstand.

5. The storage array according to any one of claims 1 to 4, characterized in that: The storage device is a spin-orbit torque magnetic storage device; The spin-orbit torque magnetic memory comprises at least a magnetic tunnel junction and a spin-orbit torque layer; The magnetic tunnel junction at least includes a free layer, a barrier layer and a reference layer stacked in sequence, and the free layer is adjacent to the spin-orbit moment layer.

6. A memory, characterized in that: The memory comprises a row address decoding module, a column address decoding module, a reading module, a column common source line, a column strobe signal line, a read bit line, a row read strobe signal line and a storage array according to any one of claims 1 to 5; The column address decoding module is connected to the control end of the first switch tube through the corresponding column selection signal line, and the row address decoding module is connected to the control end of the second switch tube through the corresponding row read selection signal line; the reading module is connected to the first end of the second switch tube through the corresponding read bit line, and the second end of the second switch tube is connected to the magnetic tunnel junction of the storage device in the corresponding row; When reading data, the row address decoding module selects the second switch tube through the signal in the corresponding row read selection signal line, and the column address decoding module selects the first switch tube through the signal in the corresponding column selection signal line. The reading module is used to read out the data in the memory through the corresponding read bit line when the first switch tube and the second switch tube are turned on at the same time.

7. The memory according to claim 6, characterized in that: The memory also includes a write module, a write bit line, and a row write word line; Each of the storage units further comprises a third switch tube, wherein three ends of the third switch tube are respectively connected to a corresponding write bit line, a first end of the corresponding storage device and a corresponding row write word line; The write module is connected to the memory device via the corresponding write bit line and the third transistor; The row address decoding module is connected to the control terminal of the third transistor through the corresponding row write word line; When data is written, the row address decoding module selects the third switch tube through the signal in the corresponding row write word line, and the column address decoding module selects the first switch tube through the signal in the corresponding column selection signal line. The write module is used to write data to the storage device through the corresponding write bit line and the corresponding column common source line when the corresponding third switch tube and the corresponding first switch tube are turned on.

8. The memory according to claim 7, characterized in that: The column address decoder is used to generate a column enable signal according to the address signal, and the column enable signal is a signal in the corresponding column enable signal line; The row address decoder is used to generate a row write signal according to the address signal, and the row write signal is a signal in the corresponding row write word line; The row address decoder is further used to generate a row read enable signal according to the address signal, and the row read enable signal is a signal in the corresponding row read enable signal line.

9. The memory according to claim 7, characterized in that: In the data reading phase, the interference voltage accumulated by the plurality of memory cells in the memory cell group on the corresponding read bit line is less than or equal to a preset interference voltage.

10. An electronic device comprising the memory according to any one of claims 6 to 9.

Citation Information

Cited By

  • Storage unit, memristor array, compute-in-memory circuit, and operation method

    WO2025119281A1