TW201525998A - Memory system - Google Patents

Abstract

A memory system of the present invention includes a memory device and a controller for controlling the memory device. The memory device is equipped with a plurality of memory cells in which data can be rewritten; a plurality of word lines connected to the plurality of memory cells; pages provided with a plurality of memory cells that are connected to the same word line; planes each provided with a plurality of pages; a memory cell array provided with a plurality of planes; a plurality of word-line drivers which apply voltage to the plurality of word lines; and a plurality of switches which are disposed on each of the planes and allocate the word-line drivers to each of the word lines.

Description

Memory system

This embodiment relates to a memory system.

Nowadays, as the use of non-volatile semiconductor memory devices (memory) expands, the capacity of memory continues to increase.

Embodiments of the present invention provide a memory system that can concurrently access pages having different page numbers located in different planes.

The memory system of this embodiment includes a memory device and a controller that controls the memory device. The memory device includes: a plurality of memory cells each holding data; a plurality of word lines connected to the plurality of memory cells; and a page having the plurality of memories connected to the same word line a plane having a plurality of pages; a memory cell array having a plurality of said planes; and a plurality of word line drivers applying a voltage to said plurality of word lines. The controller issues, to the memory device, a command to simultaneously read the first material for the specific plane and the second data for the plane different from the specific plane.

00h‧‧ directive

1‧‧‧3 dimensional laminated non-volatile semiconductor memory device (memory system)

1a‧‧‧Logical Physics Conversion Form

1b‧‧‧Write/read control unit

2‧‧‧Host

10‧‧‧BiCS flash memory

11‧‧‧ Memory Cell Array

12‧‧‧Sense Amplifier

13‧‧‧ row address buffer/row decoder

15‧‧‧Control circuit

16‧‧‧Voltage generation circuit

17‧‧‧ plane switch

17a0~17a7‧‧‧ switch

17b0~17b7‧‧‧ switch

17c0~17c7‧‧‧ switch

17n‧‧‧ switch

17o‧‧‧Switch

17p‧‧‧ switch

17q‧‧‧ switch

17r‧‧‧ switch

17s‧‧‧ switch

17t‧‧‧ switch

17u‧‧‧ switch

18‧‧‧ column address buffer

19‧‧‧Output buffer

20‧‧‧ memory controller

21‧‧‧ column decoder

21a‧‧‧Inverter

21b‧‧‧Inverter

21c‧‧‧NAND gate

21d‧‧‧Inverter

21e‧‧‧NAND gate

21f‧‧‧Inverter

21g‧‧‧NAND gate

21h‧‧‧Inverter

21i‧‧‧NAND gate

21j‧‧‧Inverter

21k‧‧‧MOS transistor

21l‧‧‧MOS transistor

21m‧‧‧MOS transistor

21n‧‧‧MOS transistor

30‧‧‧Host interface

71‧‧‧ memory cell array

72‧‧‧ line decoder

73‧‧‧ Data output buffer

74‧‧‧ Data input terminal

75‧‧‧ column decoder

76‧‧‧Control circuit

76-1‧‧‧Voltage generation circuit

77‧‧‧Control signal input terminal

78‧‧‧Source line control circuit

79‧‧‧ Well control circuit

80‧‧‧ plane switch

80a‧‧‧ switch

80b‧‧‧ switch

80c‧‧‧ switch

80d‧‧‧ switch

80h‧‧ directive

161‧‧‧Power supply

162‧‧‧CG driver

162a‧‧‧VCGSEL circuit

162b‧‧‧CGN driver

162c‧‧‧CGBG Drive/CGD Driver

162d‧‧‧CGN driver

162e‧‧‧CGU drive

162f‧‧‧VCGSEL2 circuit

162g‧‧‧CGD driver/CGBG driver

162h‧‧‧CGU drive

163‧‧‧SG driver

171a‧‧‧Flat switch CGSW

171b‧‧‧Plane switch SSGS

172a‧‧‧Flat switch CGSW

172b‧‧‧Plane switch SSGS

211‧‧‧ column decoder

212‧‧‧ column decoder

300‧‧‧NAND type flash memory

751‧‧‧ column decoder

752‧‧‧ column decoder

761‧‧‧Power supply

762‧‧‧CG driver

762a‧‧‧VCGSEL circuit

762b‧‧‧CGN driver

762c‧‧‧CGD driver

762d‧‧‧CGN driver

762e‧‧‧CGU driver

762f‧‧‧VCGSEL2 circuit

762g‧‧‧CGD driver

762h‧‧‧CGU driver

763‧‧‧SG driver

801a‧‧‧Flat switch CGSW

801b‧‧‧Flat switch SSGS

802a‧‧ plane switch CGSW

802b‧‧‧Plane Switch SGW

A‧‧‧ User Profile

ALE‧‧‧ address latch enable signal

AR‧‧‧ voltage

B‧‧ User data

B0~Bn‧‧‧ block number

Ba‧‧‧Semiconductor substrate

BG‧‧‧ back gate line

Bk‧‧‧ Block

BL‧‧‧ bit line

BLK‧‧‧ Block

BLKAD‧‧‧ block address

BLKx‧‧‧ Block

BLKy‧‧‧ Block

BR‧‧‧ voltage

BV‧‧‧ voltage

C‧‧‧User Information

C1‧‧‧ address

C2‧‧‧ address

CGBG‧‧‧ drive

CGBGI‧‧‧ signal line

CGDDB‧‧‧ drive

CGDDBI‧‧‧ signal line

CGDDBSW‧‧‧ signal

CGDDI‧‧‧ signal line

CGDDSW‧‧‧ signal

CGDDSSW‧‧‧ signal

CGDDT‧‧‧ drive

CGDDTI‧‧‧ signal line

CGDDTSW‧‧‧ signal

CGDSB‧‧‧ drive

CGDSBI‧‧‧ signal line

CGDSBSW‧‧‧ signal

CGDSI‧‧‧ signal line

CGDST‧‧‧ drive

CGDSTI‧‧‧ signal line

CGDSTSW‧‧‧ signal

CGI‧‧‧ signal line

CGN‧‧‧ drive

CGNA‧‧‧ drive

CGNB‧‧‧ drive

CGNC‧‧‧ drive

CGND‧‧‧ drive

CGSW‧‧ ‧ flat switch

CGU‧‧‧ drive

CL‧‧‧ Column

CL1‧‧‧1st semiconductor

CL2‧‧‧2nd semiconductor

CL3‧‧‧3rd semiconductor

CL4‧‧‧4th semiconductor

CLE‧‧‧Instruction Latch Enable Signal

CMD‧‧ directive

CR‧‧‧ voltage

D‧‧‧Virtual string

D1‧‧‧1st choice gate line

D2‧‧‧3rd choice gate line

E0h‧‧‧ Directive

F_DVPGHSLC‧‧‧ parameters

F_DVPGMHSLC‧‧‧ parameters

F_NLP_HSLC‧‧‧ parameters

F_VCG_HSLCV‧‧‧ parameters

F_VPGMHSLC‧‧‧ parameters

hSLC‧‧‧Information

HSLCV‧‧‧ voltage

I/O‧‧‧External input and output terminals

JP‧‧‧Link Department

LMR‧‧‧ voltage

LMV‧‧‧ voltage

MLC-Lower‧‧‧Information

MLC-Upper‧‧‧Information

MS‧‧‧NAND string

MT‧‧‧ memory cell crystal

MTr‧‧‧ memory transistor

MTr0~MTr7‧‧‧O crystal

P0~Pn‧‧‧ plane

PZ0~PZ4‧‧‧

REn‧‧‧Read enable signal

R1‧‧‧ address

R2‧‧‧ address

R3‧‧‧ address

RY/BY‧‧‧Ready/Busy signal

RZ0~RZ6‧‧‧

S2‧‧‧2nd choice gate line

SC‧‧‧U-shaped semiconductor

SGD‧‧‧汲polar selection gate line

SGDI‧‧‧ signal line

SGDTr‧‧‧汲-selective transistor

SGS‧‧‧Source side selection gate line

SGSI‧‧‧ signal line

SGSTr‧‧‧Source side selection transistor

SGSW‧‧ ‧ flat switch

SL‧‧‧ source line

SLC‧‧‧Information

SLCR‧‧‧ voltage

SLCV‧‧‧ voltage

STU‧‧‧ string unit

tR‧‧‧ reading time

USGDI‧‧‧ signal line

USGSI‧‧‧ signal line

VCELSRC‧‧‧ voltage

VCELSRCA‧‧‧ voltage

VCELSRCB‧‧‧ voltage

VCGRV‧‧‧ voltage

VCGRVA‧‧‧ voltage

VCGSEL‧‧‧ voltage

VCGSEL_AB‧‧‧ voltage

VCGSEL_CD‧‧‧ voltage

VCPWELL‧‧‧ voltage

VCPWELLA‧‧‧ voltage

VCPWELLB‧‧‧ voltage

VGP‧‧‧ voltage

VISO‧‧‧ voltage

VPASS‧‧‧ voltage

VPASS1‧‧‧ voltage

VPASS2‧‧‧ voltage

VPASSH‧‧‧ voltage

VPASSL‧‧‧ voltage

VPGM‧‧‧ voltage

VREAD‧‧‧ voltage

VREADK‧‧‧ voltage

VSGD‧‧‧Selected gate line

VSGS‧‧‧Selected gate line

VSS‧‧‧ voltage

Vth‧‧‧ threshold voltage

VUSEL1‧‧‧ voltage

VUSEL1A‧‧‧ voltage

VUSEL1B‧‧‧ voltage

VUSEL1_CD‧‧‧ voltage

VUSEL2‧‧‧ voltage

VUSEL2A‧‧‧ voltage

VUSEL2B‧‧‧ voltage

VUSEL2_CD‧‧‧ voltage

Vxxxx‧‧‧ voltage

WEn‧‧‧ write enable signal

WL‧‧‧ character line

WLD‧‧‧Dummy word line

WLDD‧‧‧ character line

WLDDB‧‧‧ character line

WLDDT‧‧‧ character line

WLDSB‧‧‧ character line

WLDST‧‧‧ character line

WLG‧‧‧ character line group

WLother‧‧‧ character line

X‧‧‧page number

Xxh‧‧ directive

Fig. 1 is a block diagram showing the circuit configuration of a three-dimensional laminated nonvolatile semiconductor memory device according to the first embodiment.

Fig. 2 is a view showing the memory cell array of the first embodiment.

Fig. 3 is a view showing one block of the p-BiCS memory of the first embodiment, connected to one; The composition of a plurality of U-shaped strings of strip bits.

Fig. 4 is a block diagram schematically showing the relationship between the driver and the plane switch of the first embodiment.

Fig. 5 is a block diagram schematically showing a CG driver of the first embodiment.

Fig. 6 is a circuit diagram of a CGN-related switch of the plane switch CCSW of the first embodiment.

Fig. 7 is a circuit diagram of a CGD-related switch of the planar switch CCSW of the first embodiment.

Fig. 8 is a circuit diagram of a decoder of the first embodiment.

Fig. 9 is a view showing a CG map at the time of a program operation of the semiconductor memory device of the first embodiment, and showing a CG map at the time of programming and reading operations of the prior art.

Fig. 10 is a view showing a CG map at the time of a reading operation of the semiconductor memory device of the first embodiment.

Fig. 11 is a view showing a CG map at the time of erasing operation of the semiconductor memory device of the first embodiment.

Fig. 12A shows the relationship between the zone signal and the CG driver when the erase operation, the program operation, and the read operation are performed.

Fig. 12B shows the relationship between the switching signal and the output signal.

Fig. 13 is a block diagram showing the basic configuration of the semiconductor memory device of the second embodiment.

Fig. 14 is a flow chart showing the operation of access integration in the second embodiment.

Fig. 15 is a view showing a case where data on a page having a different page number in a block in a different plane of the semiconductor memory device of the second embodiment is accessed in parallel.

Fig. 16 is a flow chart showing the operation of access integration in the third embodiment.

Fig. 17 is a view showing a case where data on a page having a different page number in a block in a different plane of the semiconductor memory device according to the third embodiment is accessed in parallel.

18(a)-(d) are graphs showing the distribution of the threshold value of the memory cell transistor MT.

Figure 19 shows the specific parameters of hSLC.

Fig. 20 is a view showing a reading order of the operation options of the fourth embodiment.

Fig. 21A shows the waveform of the read operation when the SLC data is read.

Fig. 21B shows the waveform of the reading operation when the MLC-Lower data is read.

Fig. 21C shows the read operation waveform when the MLC-Upper data is read.

Fig. 22A is a view showing a reading operation waveform when the SLC data of the operation option A of the fourth embodiment and the hSLC data are read.

Fig. 22B shows the read operation waveforms when the MLC-Lower data and the hSLC data are read.

Fig. 22C shows the read operation waveforms when the MLC-Upper data and the hSLC data are read.

Fig. 23A is a view showing a reading operation waveform when the SLC data and the hSLC data of the operation option B of the fourth embodiment are read.

Fig. 23B shows the read operation waveforms when the MLC-Lower data and the hSLC data are read.

Fig. 23C shows the reading action waveforms when the MLC-Upper data and the hSLC data are read.

Fig. 24 is a table showing the meaning of symbols and symbols used in the order of instructions.

Figure 25 is a diagram showing the sequence of instructions for programming the hSLC data and its internal motion waveforms.

Fig. 26 is a view showing the sequence of instructions and the internal operation waveforms when the hSLC data is read.

Fig. 27 is a view showing a specific example of the data output order.

Figure 28 is a diagram showing an example of an address used in multi-plane access.

Figure 29A shows the selection word line WL near the dummy word line WLD when the read operation is performed. A diagram of each signal in the case.

Fig. 29B is a view showing the kind of CG driver used for each word line WL and the voltage applied to the word line WL.

Fig. 30 is a block diagram schematically showing a CG driver of a sixth embodiment.

Figure 31 shows a plurality of blocks constituting a memory cell array.

Figure 32 shows a plurality of blocks constituting a memory cell array.

Figure 33 is a top plan view of a memory cell array.

Figure 34 is a representative representation of a logical block formed by a group of word lines.

Fig. 35 is a block diagram showing the basic configuration of the NAND flash memory of the eighth embodiment.

Figure 36 is a block diagram for schematically showing the relationship between the CG driver and the plane switch of the eighth embodiment.

Fig. 37 is a block diagram schematically showing the CG driver of the eighth embodiment.

Fig. 38 is a circuit diagram showing a CGD-related switch of the planar switch CCSW of the eighth embodiment.

Fig. 39 is a view showing a CG map at the time of a programming operation of the semiconductor memory device of the eighth embodiment.

Fig. 40 is a view showing a CG map at the time of a reading operation of the semiconductor memory device of the eighth embodiment.

Fig. 41 is a view showing a CG map at the time of erasing operation of the semiconductor memory device of the eighth embodiment.

Fig. 42A shows the relationship between the erase signal, the program operation, and the area signal at the time of the read operation, and the CG driver.

Fig. 42B shows the relationship between the switching signal and the output signal.

Fig. 43A is a view showing signals of a case where the word line WL is selected in the vicinity of the dummy word line WLD at the time of the read operation.

Fig. 43B is a view showing the kind of the CG driver used for each word line WL and the voltage applied to the word line WL.

Fig. 44 is a block diagram schematically showing a CG driver of a tenth embodiment.

Hereinafter, embodiments will be described with reference to the drawings. In the following description, constituent elements having substantially the same functions and configurations are denoted by the same reference numerals, and the description will be repeated only when necessary. Further, the dimensional ratio of the drawings is not limited to the illustrated ratio. In addition, each embodiment shown below exemplifies an apparatus or method for embodying the technical idea of the embodiment, and the technical idea of the embodiment is not to specify the material, shape, structure, arrangement, and the like of the components. By. The technical idea of the embodiment can be variously changed within the scope of the patent application.

(First embodiment)

<Composition of non-volatile semiconductor memory device>

Fig. 1 is a block diagram showing the circuit configuration of a three-dimensional laminated nonvolatile semiconductor memory device (memory system) according to the first embodiment.

In recent years, as a method for increasing the bit density of a NAND (Not-AND) type flash memory, a laminated type NAND flash memory in which a memory cell is laminated and a so-called BiCS (Bit- Cost Scalable) The memory of flash memory.

The three-dimensional laminated nonvolatile semiconductor memory device (memory system) 1 of the present embodiment has a BiCS flash memory (also simply referred to as a flash memory or a memory device) 10 and a memory controller 20.

Here, the BiCS flash memory 10 is provided with a memory cell array 11, a sense amplifier 12, a row address buffer/row decoder 13, a column decoder 21, a control circuit 15, a voltage generating circuit 16, a plane switch 17, The column address buffer 18 and the input/output buffer 19.

The memory cell array 11 is a three-dimensional laminated nonvolatile semiconductor memory device in which a plurality of memory cells are stacked in the vertical direction as described below. In the memory cell array 11 The memory is used, for example, to replace bad row replacement information, to determine parameters of various operation modes, or to fine-tune various voltages, and to display bad block information of bad blocks. Moreover, in one part of the memory cell array 11, the bad block information indicating the bad block generated by the day after tomorrow can also be memorized.

<Sensor Amplifier and Row Address Buffer/Row Decoder>

As shown in FIG. 1, the sense amplifier 12 is connected to the memory cell array 11 via a bit line BL. The memory cell array 11 includes a plurality of blocks BLK. For example, the data system of the memory cell transistor MT located in the same block BLK is erased. On the other hand, the reading and writing of data is performed uniformly for a plurality of memory cell transistors MT commonly connected to any of the word lines WL in any of the memory groups of any of the blocks BLK. This unit is called a "page." The sense amplifier 12 reads the data of the memory cell array 11 in page units at the time of reading, and writes the data to the memory cell array 11 in page units at the time of writing.

Also, the sense amplifier 12 is also coupled to the row address buffer/row decoder 13. The sense amplifier 12 decodes the selection signal input from the self address buffer/row decoder 13 and selects any one of the bit lines BL to drive.

The sense amplifier 12 also functions as a data latch that holds the data at the time of writing. The sense amplifier 12 of the present embodiment has a plurality of data latch circuits. A sense amplifier applied to, for example, a multi-level cell (MLC) that memorizes 2-bit data in one cell has three data latches.

The row address buffer/row decoder 13 temporarily stores the row address signal input from the memory controller 20 via the output-in buffer 19, and selects one of the selected bit lines BL according to the row address signal. The signal is output to the sense amplifier 12.

<column decoder>

The column decoder 21 decodes the column address signals input through the column address buffer 18, and selects the word line WL of the memory cell array and the selection gate lines SGD, SGS to drive. Moreover, the column decoder 21 has a portion of the block of the selected memory cell array 11 and is selected Select the part of the page.

In addition, the BiCS flash memory 10 of the present embodiment has an external input/output terminal I/O (not shown), and the data is transmitted and received by the input/output buffer 19 and the memory controller 20 via the external input/output terminal I/O. The address signals input via the external input/output terminal I/O are output to the column decoder 21 and the row address buffer/row decoder 13 via the column address buffer 18.

<control circuit>

The control circuit 15 is based on various external control signals (write enable signal WEn, read enable signal REn, command latch enable signal CLE, address latch enable signal ALE, etc.) supplied with the memory controller 20 and the instruction CMD. , sequence control of data writing and erasing, and control reading operation.

<voltage generating circuit>

The voltage generating circuit 16 is controlled by the control circuit 15 to generate various internal voltages required for the operations of writing, erasing, and reading. The voltage generating circuit 16 has a boosting circuit for generating an internal voltage higher than the power supply voltage.

<plane switch>

The planar switch 17 is connected to the control circuit 15, the voltage generating circuit 16, and the like. The plane switch 17 switches the output target of the voltage from the voltage generating circuit 16 based on the signal from the control circuit 15 or the like, and supplies it to the column decoder 21.

<memory controller>

The memory controller 20 is connected to a host (also referred to as a host device or an external device, etc.) 2 through a host interface 30. The memory controller 20 outputs an instruction or the like required for the operation of the BiCS flash memory 10, and performs reading, writing, or erasing of the BiCS flash memory 10. The memory controller 20 includes a CPU (central processing unit), a ROM (Read only memory), a RAM (Random Access Memory), or an ECC (Error Correcting Code: Error correction code) circuit.

<host>

The host 2 issues a read request or a write request to the memory controller 20 via the host interface 30. Thus, the data exchanged between the host 2 and the memory controller 20 is hereinafter referred to as user data. User data is generally managed by assigning a unique serial number called a logical address to a specific unit of 512 bytes or the like.

<Memory Cell Array>

Fig. 2 shows a memory cell array 11 of the first embodiment. In addition, in order to simplify the description, FIG. 2 sets the number of layers of the word line WL to four layers.

Fig. 2 is a perspective view showing an example of the structure of an element cell array 11 of the present embodiment. The memory cell array of the present embodiment is a p-BiCS memory in which a plurality of memory cells connected in series adjacent to each other are connected by a transistor called a catheter connection.

The memory cell array 11 has m × n (m, n is a natural number) NAND string MS. Fig. 2 shows an example of m = 6 and n = 2. Each NAND string MS is connected to a lower end of a plurality of transistors (MTr0 to MTr7) connected in series adjacent to each other, and a source side selective transistor SGSTr and a drain side selection transistor SGDTr are disposed at the upper end.

In the nonvolatile semiconductor memory device of the present embodiment, the memory transistor MTr (hereinafter referred to as a memory cell) constituting the NAND string MS is formed by laminating a plurality of conductive layers. Each NAND string MS has a U-shaped semiconductor SC, word lines WL (WL0 to WL7), a source side selection gate line SGS, and a drain side selection gate line SGD. Also, the NAND string MS has a back gate line BG.

The U-shaped semiconductor SC is formed in a U shape as viewed from the column direction. The U-shaped semiconductor SC has a pair of columnar portions CL extending in a substantially vertical direction with respect to the semiconductor substrate Ba, and a connecting portion JP formed to connect the lower ends of the pair of columnar portions CL.

The U-shaped semiconductor SC is disposed such that a straight line connecting the central axes of the pair of columnar portions CL is parallel in the row direction. Further, the U-shaped semiconductor SC is arranged in a matrix shape in a plane formed by the column direction and the row direction.

The word lines WL of the respective layers extend in parallel in the row direction. The word lines WL of the respective layers are arranged at a specific interval in the row direction, and are insulated from each other to form a line.

The gates of the memory cells (MTr0 to MTr7) disposed at the same position in the row direction and arranged in the column direction are connected to the same word line WL. Each of the word lines WL is arranged substantially vertically in the NAND string MS.

The drain side selection gate line SGD is disposed above the uppermost word line WL and extends in parallel in the column direction. The source side selection gate line SGS is also disposed above the uppermost word line WL and in parallel in the column direction, similarly to the drain side selection gate line SGD.

Further, the source side selection transistor SGSTr is connected to the common source line SL, and the drain side selection transistor SGDTr is connected to the uppermost bit line BL.

<Structure of string>

Fig. 3 is a view showing a configuration of a plurality of U-shaped strings connected to one bit line in one block of a general p-BiCS memory. The p-BiCS memory has, for example, a word line of m (m is an integer of 1 or more), and a plurality of U-shaped strings are connected to one bit line BL. A block is formed by a U-shaped string respectively connected to a plurality of bit lines BL.

Hereinafter, in each of the embodiments, a set of strings having common word lines is referred to as a physical block. Further, in each of the embodiments, the block does not mean the erase unit. The erase of the data can be performed, for example, in a string unit of the source line SL or other units.

In addition, the memory cell array 11 is composed of US Patent Application No. 12/407,403 (filed on March 19, 2009), U.S. Patent Application Serial No. 12/406,524 (filed on March 18, 2009), Patent Application No. 13/816,799 (filed on Sep. 22, 2011), and U.S. Patent Application Serial No. 12/532,030 (filed on March 23, 2009) are hereby incorporated. This application contains the entire contents of these U.S. patent applications.

<Configuration of Driver of First Embodiment>

Fig. 4 is a block diagram schematically showing the relationship between the driver and the plane switch of the first embodiment. Fig. 5 is a block diagram schematically showing a CG driver of the first embodiment.

In Fig. 4, for the sake of simplicity, the case where the memory cell array 11 has two planes will be described. Further, in the present embodiment, a case where one plane has four blocks will be described.

As shown in FIG. 4, the voltage generating circuit 16 includes a power supply 161, a CG driver (also referred to as a word line driver) 162, and an SG driver 163. The power supply 161 supplies power to the CG driver 162, the SG driver 163, and other circuits.

As shown in FIG. 5, the CG driver 162 includes CGN drivers 162b and 162d, a CGD driver 162c, and a CGU driver 162e. Further, if it is a P-BiCS, it further includes a CGBG driver 162c. The CGN driver drives the word line WL (also referred to as DataWL) that stores data in units of one.

As described below, in the programming operation of the NAND type semiconductor memory device, it is important to apply a boost from the word line WL gate to the channel of the cell in which the word line WLi (an integer of 0 or more) is not written. In order to avoid the control of channel current. Therefore, it is possible to control the voltage in the range of the unselected word line WL(i±6)~WL(i±9) in the NAND string to be optimal, and to optimally set the amount by wafer evaluation. Production. The driver used for this is a CGN driver.

It can be studied that in a NAND type semiconductor memory device in which the number of word lines WL of the NAND string is, for example, 32, a dedicated CGN driver is prepared for one word line WL for storing data, regardless of which word line is used. When WL is programmed, the optimum voltage can be applied to the word line WL before and after it. However, when the number of word lines WL of the NAND string is increased to 64 or more, if one CGN driver is prepared for one word line WL, as the number of word lines WL increases, the CGN driver also increase. As a result, there is a problem that the wafer area is increased.

In this embodiment, according to the selection of the word line WL information (area), the appropriate drive can be switched. The CGN driver of the non-selected word line WL (i ± 6) ~ WL (i ± 9) and the CGU driver described later with the word line WL other than the driving are collectively driven. This is called the decoding method of the CGN driver, whereby even if the number of WLs of the NAND string is increased to 64 to 128 or more, the number of CGN drivers can be maintained at about 16 to 24, thereby suppressing the wafer area.

Since the CGN driver switches the connection in units of division, for example, CGNA drivers <0> to <3> (concentrated as <3:0>, etc.), CGNB drivers <3:0>, CGNC drivers <3:0>, Grouped by CGND driver <3:0>. In the following, when the CGNA driver <3:0>, the CGNB driver <3:0>, the CGNC driver <3:0>, and the CGND driver <3:0> are not distinguished, it is simply referred to as a CGN driver or CGN*.

As shown in FIG. 5, the CG driver 162 of the first embodiment includes a VCGSEL circuit 162a, CGN drivers 162b and 162d (16 in total), a CGD driver 162c (four in total), a CGBG driver 162c, and a CGU driver 162e. The CG driver other than the CGU driver 162e outputs voltages of any of the voltages VCGSEL, VUSEL1, VUSEL2, and VSS. The CGU driver 162e outputs voltages VUSEL1, VUSEL2, and VSS. The VCGSEL circuit 162a, the CGN drivers 162b, 162d, the CGD driver 162c, the CGBG driver 162c, and the CGU driver 162e are controlled by control signals from the control circuit 15.

The voltage VCGSEL is the voltage selected by the VCGSEL circuit 162a. The VCGSEL circuit 162a inputs, for example, voltages VPGM and VCGRV, and the VCGSEL circuit 162a selects one based on the control signal of the control circuit 15.

The voltage VPGM is based on the voltage (cell programming voltage) applied to the selected word line WLi when the cell is programmed. The voltage VCGRV is a voltage (cell read voltage) applied to the selected word line WLi at the time of reading or program verification. The voltage VUSEL1 is the voltage VPASS1 used for channel boosting during programming. The voltage VREADK applied to the non-selected word line WL(i±1) when reading or programming verification. The voltage VUSEL2 is the voltage VPASS2 used for channel boosting during programming. read In the case of program or program verification, the voltage VREAD applied to the non-selected word line WL other than the selected word line WLi and the unselected word line WL(i±1). The voltages VCELSRC and VCPWELL are connected to the memory cell array 11.

As shown in FIG. 5, each of the CGN drivers 162b and 162d includes a circuit having a switching function for selecting a voltage applied by a plurality of word lines WL generated by the power supply 161.

The CGD driver 162c drives the word line WL (also referred to as DummyWL) that does not store data in units of one. The CGD driver has a CGDDT driver, a CGDDB driver, a CGDSB driver, and a CGDST driver. In the read operation of this embodiment, the CGDDT and CGDST drivers select, for example, the voltage VREADK output, and the CGDDB and CGDSB drivers select the VREAD output. In the following, when the CGDDT driver, the CGDDB driver, the CGDSB driver, and the CGDST driver are not distinguished, it is simply referred to as a CGD driver or CGD*.

The CGU driver 162e is an optional driver with less voltage but driving force. In the programming or reading operation, the word line WL farther from the selected word line WL is only required to be driven at the same potential, so the CGU driver is used.

In the memory cell array 11 of the present embodiment, a back gate is provided on the bottom surface of the U-string (Strings), and the CGBG driver 162c is used to drive the back gate.

The SG driver 163 supplies power to a driver such as a selection gate of the memory cell array 11.

The plane switch 17 is provided with a plane switch CCSSW and a plane switch SGSW on each plane of the memory cell array 11. More specifically, the plane switch 17 includes a plane switch CGSW 171a and a plane switch SGSW 171b corresponding to the plane <0>, and includes a plane switch CGSW 172a and a plane switch SGSW 172b corresponding to the plane <1>.

The plane switch CCSW 171a receives the zone signal ZONE_P0<3:0>, the mode signals MODE_P0<1:0>, and CGD*SW_P0 from the control circuit 15. Again, the plane switch CGSW171a from CGNA driver <3:0>, CGNB driver <3:0>, CGNC driver <3:0>, CGND driver <3:0>, CGDDT driver, CGDDB driver, CGDSB driver, CGDST driver, CGBG driver, and The CGU driver receives the signal. Further, the plane switch CCSW 171a supplies the signal received from the CG driver 162 to the column decoder 21 based on the signal from the control circuit 15. Further, the plane switch SGSW 171b supplies the SGS signal and the SGD signal received from the SG driver 163 to the column decoder 21 based on the signal from the control circuit 15.

The column decoder 21 is provided with a dedicated column decoder for each plane. More specifically, the column decoder 21 includes a column decoder 211 corresponding to the plane <0> and a column decoder 212 corresponding to the plane <1>.

The column decoder 211 receives the signals BLKADD_P0<1:0> and the signal RDEC_P0 from the control circuit 15. Further, the column decoder 211 receives signals from the plane switch CCSW 171a via the signal lines CGI<31:0>, CGDDTI, CGDDBI, CGDSBI, CGDSTI, and CGBGI. Furthermore, the column decoder 211 receives signals from the plane switches SGSW 171b via the signal lines SGSI, SGDI, USGSI, and USGDI. The column decoder 211 supplies a signal to the plane <0> based on the received signal. Further, the column decoder 212 operates in the same manner as the column decoder 211.

<Configuration of CGN-related switches of the planar switch CGSW>

Next, the configuration of the CGN-related switch of the planar switch CSWW of the first embodiment will be schematically described using FIG. Fig. 6 is a circuit diagram of a CGN-related switch of the plane switch CCSW of the first embodiment.

For example, in the present embodiment, the plane switch CCSW171a includes switches 17a0 to 17a7, 17b0 to 17b7, and 17c0 to 17c7.

CGNA<3:0> is input to one of the voltage paths of the switches 17a0, 17a4, and CGNB<3:0> is input to one of the voltage paths of the switches 17a1, 17a5. Further, CGNC<3:0> is input to one of the voltage paths of the switches 17a2 and 17a6, and the switch 17a3, Input one of the voltage paths of 17a7 to CGND<3:0>.

Further, the other end of the voltage path of the switch 17a0 is connected to the signal line CGI<3:0>, and the other end of the voltage path of the switch 17a1 is connected to the signal line CGI<7:4>. Further, the other end of the voltage path of the switch 17a2 is connected to the signal line CGI<11:8>, and the other end of the voltage path of the switch 17a3 is connected to the signal line CGI<15:12>. Furthermore, the other end of the voltage path of the switch 17a4 is connected to the signal line CGI<19:16>, and the other end of the voltage path of the switch 17a5 is connected to the signal line CGI<23:20>. Further, the other end of the voltage path of the switch 17a6 is connected to the signal line CGI<27:24>, and the other end of the voltage path of the switch 17a7 is connected to the signal line CGI<31:28>.

CGNC<3:0> is input to one of the voltage paths of the switches 17b0 and 17b4, and CGND<3:0> is input to one of the voltage paths of the switches 17b1 and 17b5. Further, CGNA<3:0> is input to one of the voltage paths of the switches 17b2 and 17b6, and CGNB<3:0> is input to one of the voltage paths of the switches 17b3 and 17b7.

Further, the other end of the voltage path of the switch 17b0 is connected to the signal line CGI<3:0>, and the other end of the voltage path of the switch 17b1 is connected to the signal line CGI<7:4>. Further, the other end of the voltage path of the switch 17b2 is connected to the signal line CGI<11:8>, and the other end of the voltage path of the switch 17b3 is connected to the signal line CGI<15:12>. Furthermore, the other end of the voltage path of the switch 17b4 is connected to the signal line CGI<19:16>, and the other end of the voltage path of the switch 17b5 is connected to the signal line CGI<23:20>. Further, the other end of the voltage path of the switch 17b6 is connected to the signal line CGI<27:24>, and the other end of the voltage path of the switch 17b7 is connected to the signal line CGI<31:28>.

The CGU is input to one of the voltage paths of the switches 17c0 to 17c7. The other end of the voltage path of the switch 17c0 is connected to the signal line CGI<3:0>, and the other end of the voltage path of the switch 17c1 is connected to the signal line CGI<7:4>. Further, the other end of the voltage path of the switch 17c2 is connected to the signal line CGI<11:8>, and the other end of the voltage path of the switch 17c3 is connected to the signal line CGI<15:12>. Furthermore, the other end of the voltage path of the switch 17c4 is connected to the letter. The line CGI<19:16>, the other end of the voltage path of the switch 17c5 is connected to the signal line CGI<23:20>. Further, the other end of the voltage path of the switch 17c6 is connected to the signal line CGI<27:24>, and the other end of the voltage path of the switch 17c7 is connected to the signal line CGI<31:28>.

Further, two kinds of signals from the control circuit 15 are input to the respective gates of the switches 17a0 to a7, 17b0 to b7, and 17c0 to c7. More specifically, signals based on the mode signals MODE<1:0> and the zone signals ZONE<2:0> are input to the gates of the switches 17a0 to a7, 17b0 to b7, and 17c0 to c7. The mode signal MODE<1:0> and the zone signal ZONE<2:0> will be described later.

<Configuration of CGD-related switches of the planar switch CGSW>

Next, the configuration of the CGD-related switch of the planar switch CCSSW of the first embodiment will be schematically explained using FIG. Fig. 7 is a circuit diagram of a CGD-related switch of the planar switch CCSW of the first embodiment.

For example, in the present embodiment, the plane switch CCSW 171a includes switches 17n, 17o, 17p, 17q, 17r, 17s, 17t, and 17u.

The CGDDT is input to one end of the voltage path of the switch 17n, the other end of the voltage path is connected to the signal line CGDDTI, and the CGDDTSW signal from the control circuit 15 is input to the gate.

The CGDDB is input to one end of the voltage path of the switch 17o, the other end of the voltage path is connected to the signal line CGDDTI, and the CGDDTSW signal from the control circuit 15 is input to the gate.

The CGDDT is input to one end of the voltage path of the switch 17p, the other end of the voltage path is connected to the signal line CGDDBI, and the CGDDBSW signal from the control circuit 15 is input to the gate.

The CGDDB is input to one end of the voltage path of the switch 17q, the other end of the voltage path is connected to the signal line CGDDBI, and the CGDDBSW letter from the control circuit 15 is input at the gate. number.

CGDST is input to one end of the voltage path of the switch 17r, and the other end of the voltage path is connected to the signal line CGDSTI, and the CGDSDSW signal from the control circuit 15 is input to the gate.

The CGDSB is input to one end of the voltage path of the switch 17s, the other end of the voltage path is connected to the signal line CGDSTI, and the CGDSDSW signal from the control circuit 15 is input to the gate.

CGDST is input to one end of the voltage path of the switch 17t, and the other end of the voltage path is connected to the signal line CGDSBI, and the CGDSBSW signal from the control circuit 15 is input to the gate.

The CGDSB is input to one end of the voltage path of the switch 17u, and the other end of the voltage path is connected to the signal line CGDSBI, and the CGDSBSW signal from the control circuit 15 is input to the gate.

<Structure of Column Decoder>

Next, the configuration of the column decoder of the first embodiment will be schematically explained using Fig. 8 . Fig. 8 is a circuit diagram of a decoder of the first embodiment.

The column decoder 21 selects the block BLK based on the block address BLKAD<0>, BLKAD<1>, and the decoding result RDEC.

That is, the MOS transistors 21k, 21l, 21m, and 21n corresponding to the block BLK included in the selected memory cell transistor MT are turned on.

For example, when the block address BLKAD<0> is "H" and the block address <1> is "L", "H" is input to the inverter 21a, and "L" is input to the inverter 21b.

Next, for the NAND gate 21c, "L" is input from the inverter 21a, and "H" is input from the inverter 21b. Since the NAND gate 21c is input with "L", the NAND gate 21c outputs "H" regardless of the RDEC. Therefore, the inverter 21d outputs "L" to the gate of the MOS transistor 21k.

Further, for the NAND gate 21e, "H" is input as the block address BLKAD<0>, and "H" is input from the inverter 21b. If RDEC is "H", the NAND gate 21c outputs "L". Therefore, the inverter 21f outputs "H" to the gate of the MOS transistor 21l. Therefore, if RDEC is "H", the block BLK1 is selected.

Further, for the NAND gate 21g, "L" is input from the inverter 21a, and "L" is input as the block address BLKAD<1>. Since the NAND gate 21g is input with "L", the NAND gate 21g outputs "H" regardless of the RDEC. Therefore, the inverter 21h outputs "L" to the gate of the MOS transistor 21m.

Further, for the NAND gate 21i, "H" is input as the block address BLKAD<0>, and "L" is input as the block address BLKAD<1>. Since the NAND gate 21i is input with "L", the NAND gate 21i outputs "H" regardless of the RDEC. Therefore, the inverter 21j outputs "L" to the gate of the MOS transistor 21n.

<Example of CG mapping>

The CG map of the first embodiment will be schematically described with reference to Figs. 9 to 11 . Fig. 9 is a view showing a CG map at the time of a programming operation of the semiconductor memory device of the first embodiment. Fig. 10 is a view showing a CG map at the time of a reading operation of the semiconductor memory device of the first embodiment. Fig. 11 is a view showing a CG map at the time of erasing operation of the semiconductor memory device of the first embodiment. In FIGS. 9 to 11, the vertical axis represents the allocation of the CG driver to the word line WL, and the horizontal axis represents the selected word line WL.

Further, in the first embodiment, the CGDST driver always applies a dedicated voltage to the word line WLDST, and the CGDSB driver always applies a dedicated voltage to the word line WLDSB. Further, in the first embodiment, the CGDDB driver always applies a dedicated voltage to the word line WLDDB, and the CGDDT driver always applies a dedicated voltage to the word line WLDDT. Furthermore, in the first embodiment, the CGBG driver always applies a dedicated voltage to the back gate BG.

<Example of CG mapping when programming action>

First, the CG mapping at the time of programming operation will be described. As shown in FIG. 9, a CG driver that applies a voltage to the word line WL is appropriately switched in accordance with the selected word line WL.

The area shown by the horizontal axis of Fig. 9 is from the control circuit 15 indicating information for connecting any of the CGN drivers or the CGU driver to each DataWL. For example, the control circuit 15 controls the plane by inputting the type of access (programming/reading/erasing, etc.) of the BiCS flash memory 10 from the memory controller 20, and the plane and page address to be accessed. The planar switch circuit sends the following MODE<1:0> and ZONE<3:0>, and indicates how each CGN and CGU driver is connected to the CGI or DataWL.

For example, when the word lines WLDST, WL0 WL WL9 are programmed, the desired voltage is applied from the CGNA driver to the word lines WL0 WL3, and the desired voltage is applied from the CGNB driver to the word lines WL4 -7. Lines WL8~11 apply the desired voltage from the CGNC driver, apply the desired voltage to the word line WL12~15 from the CGND driver, and apply the desired voltage to the word line WL16~31 from the CGU driver.

When the word lines WL10 to WL13 are programmed, switching is performed such that the CGNA driver is connected to the word lines WL16 to 19 and the CGU driver is connected to the word lines WL0 to WL3. The CGN/CGU driver-to-CGI connection determined in advance by the word line WL for programming is performed in such a manner that the combination of the connections becomes five types of the areas PZ0 to PZ4.

Each of the areas PZ0 to PZ4 is referred to as a zone at the time of programming. In order to simplify the switching circuit, the ZONE signal input from the memory controller 20 to the plane switching circuit 17 is a signal which is simply expressed as a "region" in combination with a region to be read later. At this time, the area between the programming and the reading is as close as possible to the area, and it is considered that the circuit area of the CG driver system which is roughly determined by the number of CGN drivers and the number of zones is minimized.

When the zone PZ0 is selected, the zone signal is "000" or "001", and when the zone PZ1 is selected, the zone signal is "010". When the zone PZ2 is selected, the zone signal is "011", and when the zone PZ3 is selected, the zone signal is "100". When the area PZ4 is selected, the area signals are "101", "110", and "111".

In this embodiment, it is designed to select the word line WLi (i: 0 to 31) relative to programming by using a total of 16 CGN drivers, and to select the non-selected word line WL(i+1) to be non-selected. Voltage of word line WL(i+6) (refer to D6 in the figure) or unselected word line WL(i-1) to non-selected word line WL(i-6) (refer to S6 in the figure) The voltage is optimally controlled by the CGN driver.

<Example of CG mapping when reading action>

Next, the CG mapping at the time of reading operation will be described.

When reading the NAND type semiconductor memory device, as long as a read voltage is applied to the selected word line WLi, VREADK is applied to the word line WL of the unselected word line WL(i±1), and is applied to the other word lines WL. The voltage referred to as VREAD can be used to narrow the range of word lines WL that must be controlled compared to programming, thereby reducing the number of necessary CGN drivers. Since 16 to 24 units are prepared for the programming operation, there is a CGN driver that can be applied to the CGU driver by the WL voltage at the time of reading. In the present embodiment, in the multi-plane reading operation, the CGN which is used as the CGU and the CGN driver secured by the CGN is effectively utilized in such a manner that WLs different from each other are selected for each plane.

As shown in FIG. 10, a CG driver that applies a voltage to the word line WL is appropriately switched in accordance with the selected word line WL.

As shown on the horizontal axis of Fig. 10, the areas RZ0 to RZ6 at the time of reading are set.

Specifically, when the word lines WLDST and WL0 to WL5 are read, a desired voltage is applied to the word lines WL0 to WL by a CGNA driver or a CGNC driver, and similarly, the word lines WL4 to 7 are CGNB drivers or CGND. The driver applies the desired voltage and applies the desired voltage to the word lines WL8-31 in a CGU driver. On the other hand, when the word lines WL6 to WL9 are read, switching is performed such that the CGNA driver or the CGNC driver is connected to the word lines WL8 to 11 and the CGU is connected to the word lines WL0 to WL9. The connection of the word line WL to be read is determined in advance, and the combination of the connections becomes seven types of the regions RZ0 to RZ6. Each of the regions RZ0 to RZ6 is referred to as a region at the time of reading.

When the zone RZ0 is selected, the zone signal is "000", and when the zone RZ1 is selected, the zone signal is "001". When the zone RZ2 is selected, the zone signal is "010", and when the zone RZ3 is selected, the zone signal is "011". When the zone RZ4 is selected, the zone signal is "100" or "101", and when the zone RZ5 is selected, the zone signal is "110". Also, in the case of selecting the zone RZ6, the zone signal is "111".

In this manner, in the present embodiment, at least the non-selected word line WL(i+1) can be switched by the CGN driver with respect to the selected word line WLi (i: 0 to 31) (refer to D1 in the figure), or non-selected. The voltage of the word line WL(i-1) (refer to S1 in the figure), and by using CGNA and CGNB, and CGNC and CGND as drivers for applying voltages to word lines WL of different planes, Two word lines WL are freely specified when reading in a multi-plane. In the case of, for example, 16 CGN drivers, in order to select two word lines WL for multi-plane reading, 16 CGN drivers are divided into 4 groups, and 2 of the 4 groups are assigned as 1 character. The selection of the line WL is used to allocate the remaining two groups as the selection of the other word line WL.

In the case of a 4-plane, the same multi-plane reading can be performed by dividing 16 CGN drivers into 8 groups. It can be realized without increasing the number of CGN drivers with a large circuit area.

<Example of CG mapping when erasing action>

Next, the CG mapping at the time of erasing the action will be described.

As shown in FIG. 11, during the erase operation, the CGNA driver applies voltages to the word lines WL0 to WL3 and WL16 to WL19, and the CGNB driver applies voltages to the word lines WL4 to WL7 and WL20 to WL23. Further, the CGNC driver applies voltages to the word lines WL8 to WL11 and WL24 to WL27, and the CGND driver applies voltages to the word lines WL12 to WL15 and WL28 to WL31. In addition, since this embodiment is not related to the erasing operation, detailed description is omitted.

<CG connection form>

Next, the connection table of CG will be described using FIG. 12A and FIG. 12B. Figure 12A shows the wipe In addition to actions, programming actions, zone signals when reading actions, connections from CGN/CGU drives to CGI. Figure 12B shows the relationship between the switching signal and the output signal.

As shown in FIG. 12A, when erasing, the mode signal MODE<1:0> is "00", and when programming, the mode signal MODE<1:0> is "01". At the time of reading (Read-A), the mode signals MODE<1:0> are "10", and when reading (Read-B), the mode signals MODE<1:0> are "11". In the reading (Read-A) and reading (Read-B), although the reading operation itself is substantially unchanged, the CG drivers used are different.

Figure 12B shows the connection relationship from the CGD driver to CGD*I. When the signal CGDDTSW is "0", the signal line CGDDTI is output as the output of the CGDDT driver, and when the signal CGDDDSW is "1", the signal line CGDDTI is output as the output of the CGDDB driver. Further, when the signal CGDDBSW is "0", the signal line CGDDBI is output as the output of the CGDDB driver, and when the signal CGDDBSW is "1", the signal line CGDDBI is output as the output of the CGDDT driver. When the signal CGDSDSW is "0", the signal line CGDSTI is output as the output of the CGDST driver, and when the signal CGDDSSW is "1", the signal line CGDSTI is output as the output of the CGDSB driver. Further, when the signal CGDSBSW is "0", the signal line CGDSBI is output as the output of the CGDSB driver, and when the signal CGDSBSW is "1", the signal line CGDSBI is output as the output of the CGDST driver.

<Effects of the first embodiment>

According to the first embodiment described above, the BiCS flash memory 10 includes a plurality of memory cells rewritable data and a plurality of word lines WL connected to a plurality of memory cells. Further, the BiCS flash memory 10 includes a page having a plurality of memory cells connected to the same word line WL, a plane having a plurality of pages, and a memory cell array 11 having a plurality of planes. Furthermore, the BiCS flash memory 10 is provided with: a plurality of word line drivers (CG drivers) 162 that apply voltages to the plurality of word lines WL; and a plurality of plane switches 17 disposed on each plane, and A word line driver 162 is assigned to each word Yuan line WL. When the memory controller 20 accesses a certain page existing in the BiCS flash memory 10, the flash memory is assigned a number identifying the plane to which the page belongs (referred to as a plane number), and each of the same plane is identified. The serial number of the block (called the block number) and the serial number (called the page number) of each page in the same block. Hereinafter, these are referred to as "plane number", "block number", and "page number", respectively.

However, in recent years, in NAND-type flash memories, as the miniaturization or multi-level (multi-level) development progresses, the types of voltages required have also increased. For example, when focusing on the programming of data, there are a plurality of voltages that should be applied to the unselected word lines.

For example, the voltage applied to the selected word line WLi (for example, i is 0 to 31) is VPGM at the time of programming, and is applied to the non-selected word line WL (i) adjacent to the selected word line WLi on the selected gate line SGD side. The voltage of +1) is VPASSH when programming. Further, the voltage applied to the other non-selected word line WL has VPASS, VPASSL, VGP, VISO, and the like at the time of programming.

For example, the voltage applied to the selected word line WLi is VCGRV at the time of reading, and is applied to the non-selected word line WL(i+1) or WL(i) adjacent to the selected word line WLi on the selected gate line SGD side. The voltage of -1) is VREADK when programmed. Further, the voltage applied to the other non-selected word line WL is VREAD or the like at the time of reading.

Therefore, the word line WL adjacent to the selected word line WL must be controlled by a CGN driver that can selectively output various voltages. However, CGN drivers also have the disadvantage of having a large circuit area. Therefore, the word line WL must be efficiently controlled with fewer CGN drivers.

However, in the present embodiment, the CGN driver can be allocated by the plane switch 17 for the word line WL adjacent to the selected word line WL and requiring detailed voltage adjustment. Further, for other word lines WL which can be simply applied with VPASS, a CGU driver which can select a type of voltage which can be selected by the plane switch 17 and has a smaller circuit area than a CGN driver can be allocated.

As described above, according to the present embodiment, by providing the plane switch 17 and controlling the plane switch by the area signal or the like, the CG driver can be efficiently controlled with the word line WL.

Further, even if the switching of the CG driver is not controlled by the plane switch 17 or the like, the block numbers B0, B1, ... Bn of the plurality of planes P0, P1, ..., Pn having the same page number x are generated. When the pages (P0, B0, x), (P1, B1, x), ... (Pn, Bn, x) are simultaneously accessed, the access requests may be unified and the flash memory may be The device is issued and parallel access is performed.

However, a page belonging to a different page number x, y, ... z of the blocks B0, B1, ... Bn which are present in the plurality of different planes P0, P1, ... Pn (P0) is generated. In the case of access requests of B0, x), (P1, B1, y), ... (Pn, Bn, z), pages with different page numbers sometimes belong to different word lines, so if not When the BiCS flash memory 10 described in the first embodiment is used, it is not possible to uniformly process the same.

Since the BiCS flash memory 10 of the present embodiment can allocate the CG driver to the appropriate word line WL in each plane by the plane switch 17, the page numbers in different planes can be different. The pages are accessed in parallel. That is, in the BiCS flash memory 10 of the present embodiment, due to the restriction of the word line driver (CG driver), there is no program (program) command and the Read (read) command designating the same word line WL and the same. The restrictions of the Lower/Upper page. Thereby, the performance of the NAND system can be improved.

(Second embodiment)

Next, a second embodiment will be described. In the second embodiment, multi-plane access using the semiconductor memory device described in the first embodiment will be described. In the second embodiment, the components having the same functions and configurations as those of the above-described first embodiment are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

<Configuration of Semiconductor Memory Device According to Second Embodiment>

First, the basic configuration of the semiconductor memory device of the second embodiment will be schematically described using FIG. Fig. 13 is a block diagram showing the basic configuration of the semiconductor memory device of the second embodiment.

As shown in FIG. 13, the memory controller 20 further includes a logical-physical conversion table 1a, and The write/read control unit 1b.

The logical-physical conversion table 1a is a table for maintaining information on which physical memory cell location (physical address) of the user data stored in the BiCS flash memory 10 having the logical address supplied from the host 2 through the host interface 30. . When the host 2 supplies the logical address, the memory controller 20 uses the logical physical conversion table 1a to derive the physical address corresponding to the received logical address.

The write/read control unit 1b can be realized by a hardware such as a CPU (not shown), or can be realized by a software that operates on a CPU or a RAM. The detailed operation of the write/read control unit 1b will be described later.

<Access integration operation of the second embodiment>

Next, the access integration operation of the second embodiment will be described with reference to Figs. 14 and 15 . Fig. 14 is a flow chart showing the operation of access integration in the second embodiment. Fig. 15 is a view showing a case where data parallel accesses of pages having different page numbers stored in blocks in different planes of the BiCS flash memory 10 of the second embodiment are illustrated. Here, for simplicity, the kth page of the jth block of the i-th plane is represented as a page (i, j, k).

[Step S1001]

The memory controller 20 receives an access request (access command) from the host 2. Here, the access request includes a read request (read command) and a write request (write command).

[Step S1002]

After receiving the access request from the host 2, the memory controller 20 starts the processing of the received access request at the point in time at which processing can be started.

[Step S1003]

The write/read control unit 1b determines whether or not there is an access request for a plurality of processing targets. For example, the memory controller 20 includes, for example, a command queue area (not shown). The command queue area maintains instructions received from the host 2. The write/read control unit 1b can refer to the access request held in the command queue area to determine whether or not there are a plurality of access requests.

[Step S1004]

In the case where the write/read control unit 1b determines that there are a plurality of access requests to be processed in the step S1003, the write/read control unit 1b determines whether or not a plurality of access requests can be integrated. If the plurality of access requests of the processing object are for pages in different planes of the same BiCS flash memory 10, it is determined that the access requests can be integrated, and if not, the determination is not integrated. More specifically, regarding the plurality of access requests to be processed, the write/read control unit 1b refers to the physical address of the access target (derived in S1002). When a page on a plane different from the physical address of the access target of the plurality of access requests is specified, and the access content is the same as the read request to the write request, it is determined that the plurality of access requests can be integrated. On the other hand, when the above conditions are not satisfied, it is determined that the integration is not possible. Here, the integrated access request refers to the sequential issuance of the instructions to the BiCS flash memory, and simultaneously performs the reading operation to the programming operation on the plurality of data having the access request.

[Step S1005]

In step S1004, when the write/read control unit 1b determines that a plurality of access requests can be integrated, the write/read control unit 1b integrates a plurality of access requests.

[Step S1006]

In step S1003, when the write/read control unit 1b determines that the access request to be processed is not plural (that is, when the access request of the processing target is determined to be one), the write/read control unit 1b accesses the BiCS flash memory 10 based on the access request of the processing object.

In step S1004, when the write/read control unit 1b determines that a plurality of access requests cannot be integrated, the write/read control unit 1b sequentially performs the BiCS flash memory 10 based on each access request. access.

When the write/read control unit 1b integrates a plurality of access requests, the write/read control unit 1b performs access to the BiCS flash memory 10 based on the integrated access request (referred to as Parallel access or multi-plane access).

Access to the BiCS flash memory 10 based on the integrated access request is illustrated using FIG.

As described above, the write/read control unit 1b integrates a plurality of access requests when a plurality of access requests of the processing target are pages in different planes of the same BiCS flash memory 10, and is based on The integrated access request performs access to the BiCS flash memory 10. Figure 15 shows the access to the BiCS flash memory 10 based on the integrated access request.

Figure 15 is a diagram showing an access request for the user profile A existing on the page (0, 1, 1), an access request for the user profile B existing on the page (1, 2, 0), and An access request to the user profile C existing on the page (n, 1, 2), and access based on the integrated access request.

As shown in FIG. 15, in the semiconductor memory device of the second embodiment, pages (0, 0, 1), (1, 2, 0), ... (n, 1) having different page numbers can be stored. , 2) User data A, B, C are accessed in parallel. In addition, the parallel access means that the memory cell stored in the user data A, B, and C having the access request is simultaneously read to the programming operation.

<Operation and Effect of Second Embodiment>

According to the second embodiment described above, the BiCS flash memory that can be accessed in parallel in the pages of different page numbers in different planes described in the first embodiment is integrated, and the pages stored in different planes are integrated. Access requests for data on pages with different serial numbers and access in parallel.

In this way, by integrating a plurality of access requests, it is not necessary to perform data access successively, thereby improving the processing capability of data access to the BiCS flash memory.

(Third embodiment)

Next, a third embodiment will be described. In the third embodiment, multi-plane access is explained. Other examples. In the third embodiment, the components having the same functions and configurations as those of the above-described second embodiment are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

<Access integration operation of the third embodiment>

Next, the access integration operation of the third embodiment will be described with reference to Figs. 16 and 17 . Fig. 16 is a flow chart showing the operation of access integration in the third embodiment. Fig. 17 is a view showing a case where data parallel accesses of pages having different page numbers stored in blocks in different planes of the BiCS flash memory 10 of the third embodiment are illustrated. Generally, when the memory controller 20 receives a request from the host 2 to access the user data, some data corresponding to the user data but not the user data itself is used as the secondary information. Hereinafter, the data that the memory controller 20 uses for internally managing user data is referred to as metadata.

In the third embodiment, the memory controller 20 stores meta-data corresponding to a certain user data in different planes in the same flash memory device. At this time, the page number and the page number of the page for storing the user data and the metadata corresponding thereto may be different. Further, in the third embodiment, the user data is stored in the plane 0 to n-1, and the metadata is stored in the plane n, but the present invention is not limited thereto.

Moreover, when determining the storage location (physical address) of the meta-data corresponding to a user profile, the storage location of the metadata may be configured as an attribute that can be easily stored according to the storage location (physical address) of the corresponding user profile. Calculation. With such a configuration, the advantage of easily obtaining the position of the metadata corresponding to the user data can be obtained.

[Step S2001]

The memory controller 20 receives the read request from the host 2 (refer to step S1001 of FIG. 14).

[Step S2002]

The memory controller 20 starts receiving after receiving a read request from the host 2. The processing of the read request is completed (refer to step S1002 of FIG. 14).

[Step S2003]

After starting the access request processing from the host 2, the memory controller 20 acquires the storage location of the metadata corresponding to the user data having the access request.

[Step S2004]

The write/read control unit 1b issues a command sequence for accessing the user data having an access request from the host 2 and the metadata corresponding to the user data in parallel to the BiCS flash memory 10 (refer to FIG. 14). Step S1006).

Thereby, the memory controller 20 can perform access to the user data having an access request from the host 2 and the metadata corresponding to the user data in parallel.

As shown in FIG. 17, the user data A stored in the page (0, 0, 1) is associated with the metadata A stored in the page (n, 1, 2) by the memory controller 20. Similarly, the user data B stored in the page (1, 2, 0) is associated with the metadata B stored in the page (n, 1, 1) by the write/read control unit 1b.

As shown in FIG. 17, the write/read control unit 1b can access the user data and the metadata group in parallel. This is because the memory controller 20 stores the user data A, B and the metadata A and B corresponding thereto in different planes in the same BiCS flash memory 10.

<Effects of the third embodiment>

According to the third embodiment described above, the user data and the metadata associated with the user data are stored in different planes on the same BiCS flash memory 10. When the memory controller 20 accesses the user data and the metadata associated with the user profile, the BiCS flash memory 10 issues an instruction sequence for accessing the two data in parallel. Thereby, the user data and the metadata associated with the user profile can be accessed in parallel.

By accessing the user data in parallel and the metadata associated with the user data, the data access can be performed one by one, thereby improving the processing capability.

(Fourth embodiment)

Next, a fourth embodiment will be described. In the fourth embodiment, the programming operation and the reading operation of the semiconductor memory device described in the first embodiment, the second embodiment, and the third embodiment will be described. In the fourth embodiment, the components having the same functions and configurations as those of the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

<Example of programming method of data>

Hereinafter, a case where a charge is injected into the charge accumulating layer to raise the threshold voltage of the memory cell MT is referred to as "x0" programming, "00" programming, "10" programming, "0" programming, or the like. On the other hand, a case where the charge is not injected into the charge accumulating layer so as not to change the threshold voltage (in other words, it is suppressed to a level that does not cause the data to be converted to other levels) is called "x1". "Programming, "11" programming, "1" programming, etc.

An example of a method of programming data in a memory according to the present embodiment will be described using FIG. However, for simplification of the description, the following describes a case of a 4-value (4-levels) or a 2-value (2-levels) NAND-type flash memory. Also, the same is true for other multi-bit NAND type flash memories.

Fig. 18 is a graph showing the distribution of the threshold value of the memory cell transistor MT. When the memory cell transistor MT can maintain the 4-value data (2-bit data), the memory cell transistor MT can maintain the "11", "01", "00", and "10" in ascending order of the threshold voltage Vth. "4 kinds of information.

<Example of MLC programming method>

Fig. 18(a) is a graph showing the distribution of the threshold value of the memory cell transistor MT, showing the change in the threshold distribution after the programming of the lower page in the MLC programming. Fig. 18(b) is a graph showing the distribution of the threshold value of the memory cell transistor MT, showing the distribution of the threshold value of the upper page of the MLC (multi level cell) programming. Change.

As shown in Fig. 18 (a) and (b), the data is uniformly written to one page. Moreover, the data is written in units of one bit of two bits. At this time, as shown in the figure, the data of the lower order bit in the 2-bit element is first written, and then the data of the upper order bit is written. In the case of "0" programming of the lower order bit, coarse programming is performed. Then, when programming the upper-order bit, programming is performed in such a way that the threshold value becomes higher than BV when "00" is programmed, and when the "10" programming is performed, the threshold is made. Programming becomes faster than CV.

<Example of SLC programming method>

Fig. 18(c) is a graph showing the distribution of the threshold value of the memory cell transistor MT, showing the change in the threshold distribution after programming of the SLC (Single level cell).

As shown in Fig. 18(c), coarser programming is performed compared to MLC programming. Next, when "0" programming is performed, programming is performed in such a manner that its threshold becomes higher than SLCV.

Further, the SLC data is used, for example, for metadata.

<Example of hSLC programming method>

Fig. 18(d) is a graph showing the distribution of the threshold value of the memory cell transistor MT, showing the change in the threshold distribution after programming with hSLC (higher Single level cell).

As shown in Fig. 18(d), when "0" programming is performed, higher programming is performed than "0" of SLC programming. More specifically, when "0" programming is performed, programming is performed in such a manner that its threshold becomes higher than hSLCV. During hSLC programming, the hSLC-specific parameters shown in Figure 19 are used for programming. The parameter F_VPGMHSLC defines the parameters of the initial voltage VPGM in hSLC programming. The parameter F_DVPGMHSLC is a parameter that defines the amount of increase in VPGM in hSLC programming. The parameter F_VCG_HSLCV defines the parameters for verifying the level in hSLC programming. The parameter F_NLP_HSLC defines the maximum number of loops in hSLC programming. The parameters. In addition, the parameters are pre-stored in a portion of the memory array 11 and are transmitted to the registers in the control circuit 15 when the power is turned on.

Further, the hSLC data is used, for example, in metadata. The "0" of the hSLC data is described as an example of BV or more as MLC in advance.

<reading order>

Next, the operation options of the fourth embodiment will be described using FIG. Fig. 20 is a view showing a reading order of the operation options of the fourth embodiment. Here, for simplification, two planes of plane 0 and plane 1 are extracted.

<Action Option A>

First, the action option A will be explained. The hSLC data is dedicated to the threshold voltage of the memory cell to be more than BV of the MLC. Therefore, the action option A is to read the SLC data and the MLC Lower/Upper data on the one hand, and simultaneously read the hSLC data on the other hand. This action option A prevents the read time from increasing.

When reading the SLC data of plane 0 and the hSLC data of plane 1, the data of both can be read by applying SLCR to the selected word line WL.

When reading the MLC-Lower of plane 0 (which has not been subjected to Upper programming) data and the hSLC data of plane 1, the hSLC data can be read by applying BR to the selected word line WL, after which, by The LMR is applied to the selected word line WL to read the MLC-Lower data.

In the case of reading the MLC-Upper of plane 0 (the upper programming has not been performed) and the hSLC data of plane 1, the hSLC data can be read by applying AR to the selected word line WL, and thereafter, by The CR is applied to the selected word line WL to read the MLC-Upper data.

In the case of reading the MLC-Lower (Upper Programming) data of plane 0 and the hSLC data of plane 1, the data of both can be read by applying BR to the selected word line WL.

In the case of reading the MLC-Upper (Upper Programming) data of plane 0 and the hSLC data of plane 1, the hSLC data can be read by applying AR to the selected word line WL, and thereafter, by The CR is applied to the selected word line WL to read the MLC-Upper data.

In the action option A, without the reading time tR, the data of the plane 0 (user data) and the data of the plane 1 (meta data) can be read.

Also, when this option is selected, F_HSLC_MODE is set to "1".

<Action Option B>

Next, the action option B will be explained. For the hSLC data, the threshold voltage of the memory cell is equal to or greater than the LMV of the MLC by SLC instruction and SLCV or more. Therefore, the action option B is to read the SLC data and the MLC Lower/Upper data on the one hand, and also to read the hSLC data at the same time. The action option B is that the write level of the SLC is substantially the same as the write level of the hSLC, so the W/E times of the hSLC data are approximately the same as the W/E times of the SLC data. When the programming level of the hSLC data is exactly the same as the programming level of the SLC, the SLC programming instruction can also be used when programming the hSLC data, and when the writing level is finely adjusted, only the dedicated instruction of the hSLC data is used. Just fine.

In the case of reading the SLC data of plane 0 and the hSLC data of plane 1, the data of both can be read by applying SLCR to the selected word line WL.

The hSLC and MLC-Lower data can be read by applying LMR to the selected word line WL when reading the MLC-Lower of plane 0 (which has not been subjected to Upper programming) data and the hSLC data of plane 1.

When reading the MLC-Upper of Plane 0 (without Upper Programming) data and the hSLC data of Plane 1, by applying AR to the selected word line WL, the hSLC data can be read, and thereafter, by selecting The character line WL applies CR to read the MLC-Upper data.

MLC-Lower (Upper Programming) data for reading plane 0, and plane 1 In the case of the hSLC data, the MLC-Lower data can be read by applying LMR to the selected word line WL, after which the hSLC data can be read by applying BR to the selected word line WL.

In the case of reading the MLC-Upper (Upper Programming) data of plane 0 and the hSLC data of plane 1, by applying AR to the selected word line WL, the hSLC data can be read, and thereafter, by selecting The character line WL applies CR to read the MLC-Upper data.

In action option B, hSLC can be written above SLCV, which is the same tPROG/reliability (W/E times) as SLC. Moreover, in the action option B, since the programming of the hSLC can be performed by using the programming instruction of the SLC, multi-plane programming with the SLC data can be realized (the word line WL during programming must be the same in all planes). Also, in the case of the action option B, when the hSLC programming instruction is used to program the hSLC, the SLC can change the Vth distribution (threshold distribution). Also, when this option is selected, F_HSLC_MODE is set to "0".

<action waveform>

Next, the operation waveforms at the time of the reading operation of the present embodiment will be described with reference to FIGS. 21A, 21B, 21C, 22A, 22B, 22C, 23A, 23B, and 23C.

Fig. 21A shows the read operation waveform when the SLC data is read, Fig. 21B shows the read operation waveform when the MLC-Lower data is read, and Fig. 21C shows the read operation waveform when the MLC-Upper data is read.

Fig. 22A shows the read operation waveforms when the SLC data and the hSLC data are read in the operation option A of the embodiment, and Fig. 22B shows the read operation waveform when the MLC-Lower data and the hSLC data are read, and Fig. 22C shows the MLC. -Upper data and read motion waveforms when reading hSLC data.

FIG. 23A shows the read operation waveforms when the SLC data and the hSLC data are read in the operation option B of the embodiment, and FIG. 23B shows the read operation waveform when the MLC-Lower data and the hSLC data are read, and FIG. 23C shows the MLC. -Upper data and read motion waveforms when reading hSLC data.

As shown in FIG. 21A, at the time of reading of the SLC data, a voltage SLCR is applied to the selected word line WLn, and a voltage VREADK (VREADK> is applied to the unselected word lines WL(n+1), WL(n-1). SLCR), and applies voltage VREAD (VREADK>VREAD>SLCR) to other non-selected word lines WL (referred to as WLother, etc.).

As shown in FIG. 21B, at the time of reading the MLC-Lower data, a voltage BR is applied to the selected word line WLn, and a voltage VREADK is applied to the unselected word lines WL(n+1), WL(n-1) ( VREADK>BR), and applies a voltage VREAD (VREADK>VREAD>BR) to the other non-selected word line WLother. When the upper programming is not performed (refer to the dotted line portion), the voltage LMR (BR>LMR) is applied to the selected word line WLn, and the non-selected word lines WL(n+1), WL(n-1) are applied. The voltage VREADK (VREADK>BR>LMR) is applied to the other non-selected word line WLother by applying a voltage VREAD (VREADK>VREAD>BR>LMR).

As shown in FIG. 21C, at the time of reading of the MLC-Upper data, a voltage AR is applied to the selected word line WLn, and thereafter a voltage CR (CR>AR) is applied to the selected word line WLn, and the non-selected word line WL is applied. (n+1), WL(n-1) applies a voltage VREADK (VREADK>CR>AR), and applies a voltage VREAD (VREADK>VREAD>CR>AR) to the other non-selected word line WLother.

Further, as shown in FIG. 21A and FIG. 22A, the operation waveforms at the time of reading the SLC data and the hSLC data in the operation option A are the same as those at the time of reading the SLC data.

Further, as shown in FIG. 21B and FIG. 22B, the action waveforms at the time of reading the MLC-Lower data and the hSLC data in the operation option A are the same as those at the time of reading the MLC-Lower data.

Further, as shown in FIG. 21C and FIG. 22C, the action waveforms at the time of reading the MLC-Upper data and the hSLC data in the operation option A are the same as those when the MLC-Upper data is read.

Moreover, as shown in FIG. 21A and FIG. 23A, the SLC data and the hSLC data in the action option B are shown. The waveform of the action at the time of reading is the same as the waveform of the operation when the SLC data is read.

As shown in FIG. 23B, when reading the MLC-Lower data and the hSLC data, a voltage LMR is applied to the selected word line WLn, and thereafter a voltage BR (BR>LMR) is applied to the selected word line WLn, and the non-selected word is applied. The voltages VREADK (VREADK>BR>LMR) are applied to the WL(n+1) and WL(n-1), and the voltage VREAD (VREADK>VREAD>BR>LMR) is applied to the other non-selected word line WLother. Since the action reads the hSLC data, the voltage LMR is required, and the frequency of the read operation is higher after the upper program is executed than before the upper program is executed. Since the frequency of the two readings of the voltage LMR and the voltage BR is relatively high, the voltage LMR and the voltage BR are continuously applied in sequence, thereby reducing the discharge of the unselected word line WL and the reset time of the sense amplifier not shown. High speed. Alternatively, the method of Fig. 22B which is not continuously applied may be employed.

Further, as shown in FIG. 21C and FIG. 23C, the action waveforms at the time of reading the MLC-Upper data and the hSLC data in the operation option B are the same as those at the time of reading the MLC-Upper data.

<Example of programming sequence>

Next, the programming sequence of the fourth embodiment will be described with reference to Figs. 24 and 25 .

Figure 24 is a table showing the meaning of the order of the instructions and the meaning of the symbols used to display the order of the instructions.

Figure 25 is a diagram showing the sequence of instructions for programming the hSLC data and its internal motion waveforms. Here, an example in which the hSLC data is programmed for the plane 0, the block BLKX, and the selected word line WLn will be described. In addition, the R/B shown in the figure indicates the ready/busy signal line between the memory controller 20 and the BiCS flash memory 10.

As shown in FIG. 25, when the hSLC data is programmed, the memory controller 20 issues a xx instruction and an 80h instruction to the BiCS flash memory 10. Next, the memory controller 20 issues a specific command or address, write data, and the like to the BiCS flash memory 10.

As shown in FIG. 25, the BiCS flash memory 10 is input from the memory controller 20. When the programming instruction of the hSLC data is commanded, the memory cell array 11 is programmed.

As shown in FIG. 25, the voltage VPGM is applied to the selected word line WLn at the time of the first (LOOP#=1) programming, and the various voltages Vxxxx (VPGM>Vxxxx) as described above are applied to the other non-selected word lines WLother. In addition, the voltage VPGM can be set by the parameter F_VPGMHSLC.

Next, the first verification operation is performed to determine whether or not programming is normally performed. In the verify operation, a voltage HSLCV is applied to the selected word line WLn, and a voltage VREADK (VREADK>HSLCV) is applied to the non-selected word lines WL(n+1), WL(n-1), and other non-selected words are applied. The line WLother applies a voltage VREAD (VREADK>VREAD>HSLCV). In addition, the voltage HSLCV can be set by the parameter F_VCG_HSLCV.

Here, when the programming is not completed by the first programming operation, the second programming operation is performed.

At the second (LOOP#=2) programming, a voltage VPGM higher than the voltage VPGM at the time of the first programming operation is applied to the selected word line WLn, and the other non-selected word line WLother is applied as above. The various voltages Vxxxx (VPGM>Vxxxx) are described. In addition, the voltage VPGM during the programming operation is increased by the voltage VPGM from the last programming operation by the parameter F_DVPGHSLC.

Next, in the same manner as in the case of the first verification operation, the second verification operation is performed to determine whether or not the programming is normally performed.

As shown in FIG. 25, when the program operation is not performed until the maximum number of times (LOOP#=MAX) verification has not been turned on, the state register setting of the unillustrated BiCS flash memory 10 fails and ends. Programming action. In addition, the number of loops can be set by the parameter F_NLP_HSLC.

<Example of reading order>

Next, the reading order of the fourth embodiment will be described with reference to Fig. 26 .

Figure 26 shows the sequence of instructions and the waveform of the action when programming the hSLC data. Picture. Here, the HSLC data held in the plane 0, the block BLKx, and the selected word line WLn, and the MLC-Upper data held in the plane 1, the block BLKy, and the selected word line WLm are simultaneously read. Description. In this way, the reading operation performed on different planes at the same time is called multi-plane reading or the like. This multi-plane read has the same meaning as the parallel access described above. In addition, the R/B shown in the figure indicates the ready/busy signal line between the memory controller 20 and the BiCS flash memory 10.

As shown in FIG. 26, in the case of reading the hSLC data, the memory controller 20 issues a xx instruction and a 00h instruction to the BiCS flash memory 10. Then, the memory controller 20 issues a specific instruction or address to the BiCS flash memory 10. Further, the memory controller 20 issues a 00h command to the BiCS flash memory 10 for reading the MLC-Upper data. Then, the memory controller 20 issues a specific instruction or address to the BiCS flash memory 10.

As shown in FIG. 26, when the BiCS flash memory 10 inputs each command from the memory controller 20, the read operation is performed from the memory cell array 11.

In plane 0, block BLKx, page hSLC, a voltage AR is applied to the selected word line WLn, after which a voltage CR (CR > AR) is applied to the selected word line WLn, and the non-selected word line WL (n+) is applied. 1), WL(n-1) applies a voltage VREADK (VREADK>CR>AR), and applies a voltage VREAD (VREADK>VREAD>CR>AR) to the other non-selected word line WLother. In addition, the hSLC data is read by applying a voltage AR to the selected word line WL.

In the plane 1, the block BLKy, the page MLC-Upper, a voltage AR is applied to the selected word line WLm, and thereafter, a voltage CR (CR > AR) is applied to the selected word line WLn, and the non-selected word line WL is applied ( m+1), WL(m-1) applies a voltage VREADK (VREADK>CR>AR), and applies a voltage VREAD (VREADK>VREAD>CR>AR) to the other non-selected word line WLother.

In addition, access to plane 0 and plane 1 can be performed using the BiCS described in the first embodiment. The flash memory 10 is simultaneously accessed.

In this way, hSLC data and MLC-Upper data can be continuously read.

<Specific example of data output order>

Next, a specific example of the reading order and the data output order will be briefly described using FIG. 27 and FIG. 28. Fig. 27 is a view showing a specific example of the data output order. Figure 28 is a diagram showing an example of an address used in multi-plane access. In addition, the R/B shown in the figure indicates the ready/busy signal line between the memory controller 20 and the BiCS flash memory 10.

As shown in FIG. 27, when the hSLC data held by the plane 0 is outputted, the memory controller 20 issues a 00h command to the BiCS flash memory 10. Then, the memory controller 20 issues the addresses C1, C2, R1, R2, R3, and the instruction E0h to the BiCS flash memory 10. Thereby, the hSLC data (R-Data) is output from the BiCS flash memory 10.

As shown in FIG. 27, when the MLC data held by the plane 1 is output as data, the memory controller 20 issues a 00h command to the BiCS flash memory 10. Then, the memory controller 20 issues the addresses C1, C2, R1, R2, R3, and the instruction E0h to the BiCS flash memory 10. Thereby, the MLC data (R-Data) is output from the BiCS flash memory 10.

As shown in Fig. 28, in the present embodiment, as an example, the physical address on the flash memory is represented by an address of 8 bits × 5 cycles.

For example, in the present embodiment, the bits of R1-1 to R1-5 are used when representing the serial number of the word line WL, and the bits of R2-3 to R2-4 are used when the plane number is indicated.

In the present application, when multi-plane access is performed on the hSLC data and the MLC data, there is a possibility that the word line WL number and the plane number of the hSLC data are different from the character line WL number and the plane number of the MLC data. Sex.

<Effects of the fourth embodiment>

According to the fourth embodiment described above, the MLC data or the SLC data is used as the user data, and the hSLC data or the SLC data is used as the metadata. Then, the normal reading (SLC, MLC-Lower/Upper data) can be performed on a certain plane while reading on other planes. hSLC data (metadata).

For example, regardless of whether the SLC, MLC-Lower/Upper data is read, as long as it is stored in a plane different from the plane in which the reading is performed, it can be used with SLC, MLC-Lower/Upper data. At the same time as the reading operation of one, the reading operation of the hSLC data is performed.

Further, if the MLC-Lower/Upper data, or the SLC data, or the hSLC data are mutually different, different word lines WL can be selected in each plane in the multi-plane operation.

As described above, by simultaneously accessing a plurality of planes in parallel, data access can be performed sequentially, thereby improving processing capability and improving the performance of the NAND system.

(Fifth Embodiment)

Next, a fifth embodiment will be described. In the fifth embodiment, the reading operation in the vicinity of the dummy word line will be described. In the fifth embodiment, the components having the same functions and configurations as those of the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

The reading operation when the word line WL is selected in the vicinity of the dummy word line WLD will be described with reference to FIGS. 29A and 29B. 29A shows a zone signal ZONE<3:0>, a mode signal MODE<1:0>, a CGDDTSW signal, a CGDDBSW signal, and a CGDSDSW signal in the case where the word line WL is selected near the dummy word line WLD in the read operation. And the map of the CGDSSSW signal. Fig. 29B is a view showing the kind of the CG driver used for each word line WL and the voltage applied to the word line WL. In addition, here is a simplification, and attention is paid to plane 0 and plane 1, indicating the case where hSLC data is read from plane 1.

As shown in FIG. 29A and FIG. 29B, in the case of the plane 0, when the word line WL is selected, for example, in the case of the word line WL31 adjacent to the dummy word line WLDDT, the area signal ZONE<3:0> is "111". The mode signal MODE<1:0> is "10", the CGDDTSW signal is "0", the CGDDBSW signal is "0", the CGDSDSW signal is "1", and the CGDSBSW signal is "0".

For the selected word line WL31, the voltage VCGRV is applied by the CGNB<3> driver, and the voltage VREADK (VREADK>VCGRV) is applied to the non-selected word line WL30 by the CGNB<2> driver. Next, for the dummy select word line WLDDT, a voltage VREADK (VREADK > VCGRV) is applied by the CGDDT driver.

As shown in FIG. 29A and FIG. 29B, in the case of the plane 1, when the word line WL is selected, for example, in the case of the word line WL15 adjacent to the dummy word line WLDSB, the area signal ZONE<3:0> is "011". The mode signal MODE<1:0> is "11", the CGDDTSW signal is "1", the CGDDBSW signal is "0", the CGDSDSW signal is "1", and the CGDSSSW signal is "1".

For the selected word line WL15, the voltage VCGRV is applied by the CGND<3> driver, and the voltage VREADK (VREADK>VCGRV) is applied to the non-selected word line WL14 by the CGND<2> driver. Next, for the dummy select word line WLDSB, a voltage VREADK (VREADK > VCGRV) is applied by the CGDST driver.

(Sixth embodiment)

Next, a sixth embodiment will be described. In the sixth embodiment, a CG driver and a power supply device different from the CG driver and the power supply device described in the first embodiment will be described. In the sixth embodiment, the components having substantially the same functions and configurations as those of the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

As shown in Fig. 30, the power supply 161 and the CG driver 162 of the sixth embodiment divide the power supply into a plane A and a plane B. As shown in FIG. 30, the CG driver 162 of the sixth embodiment includes a VCGSEL circuit 162a, CGN drivers 162b and 162d (16 in total), a CGD driver 162c (four in total), a CGBG driver 162c, a CGU driver 162e, and a VCGSEL2 circuit 162f. The CGD driver 162g (four in total), the CGBG driver 162g, and the CGU driver 162h.

The VCGSEL circuit 162a outputs a voltage according to a control signal from the control circuit 15. VPGM or VCGRVA is used as the voltage VCGSEL_AB.

The CGN driver 162b, the CGD driver 162c, and the CGBG driver 162c output the voltages of any of the voltages VCGSEL_AB, VUSEL1A, VUSEL2A, and VSS to the plane A based on the control signal from the control circuit 15.

The CGU driver 162e outputs the voltages of any of the voltages VUSEL1A, VUSEL2A, and VSS to the plane A based on the control signal from the control circuit 15.

The VCGSEL2 circuit 162f outputs voltages VPGM, VCGRVA, and VCGRVB as voltages VCGSEL_CD based on control signals from the control circuit 15. Further, the VCGSEL2 circuit 162f outputs the voltages VUSEL1A and VUSEL1B as the voltage VUSEL1_CD based on the control signal from the control circuit 15. Further, the VCGSEL2 circuit 162f outputs the voltages VUSEL2A and VUSEL2B as the voltage VUSEL2_CD based on the control signal from the control circuit 15.

The CGN driver 162d, the CGD driver 162g, and the CGBG driver 162g output the voltages of any of the voltages VCGSEL_CD, VUSEL1_CD, VUSEL2_CD, and VSS to the plane B based on the control signal from the control circuit 15.

The CGU driver 162h outputs the voltages of any of the voltages VUSEL1_CD, VUSEL2_CD, and VSS to the plane B based on the control signal from the control circuit 15. The voltages VCELSRCA and VCPWELLA are connected to the memory cell array 11 of the plane A. The voltages VCELSRCB and VCPWELLB are connected to the memory cell array 11 of the plane B. In addition, the plane A and the plane B may be any plane.

<Effects of the sixth embodiment>

According to the sixth embodiment, the power supply unit 161 of the sixth embodiment has two voltage systems for two planes, so that voltage can be applied to both planes at the same time as compared with the power supply unit 161 of the first embodiment. The CG driver is constructed. Therefore, for example, the above MLC data, and hSLC data, or SLC and hSLC data can be simultaneously read.

(Seventh embodiment)

Next, a seventh embodiment will be described. In the seventh embodiment, a memory cell array different from the memory cell array 11 described in the first embodiment will be described. In the seventh embodiment, functions and configurations that are substantially the same as those in the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

<Composition of string unit STU>

In the BiCS flash memory, the columnar semiconductor constituting the channel is formed in the opening portion having a relatively large aspect. With the development of the miniaturization of the BiCS flash memory, it is required to shorten the distance (distance) between the openings, and it has been studied to arrange the openings in a zigzag shape.

31 to 34 show the configuration of the seventh embodiment. The seventh embodiment is different from the configuration of the string unit of the first to fifth embodiments.

Figure 31 shows a plurality of blocks Bk-1, Bk, Bk+1 constituting a memory cell array. A plurality of blocks Bk-1, Bk, and Bk+1 are arranged along a plurality of bit lines BL. Each of the plurality of blocks Bk-1, Bk, and Bk+1 has a plurality of string units STUs arranged in the bit line direction. Each string unit STU is composed of, for example, three selection gate lines and three word line groups.

Further, a dummy string D is disposed between the blocks. Since each of the plurality of blocks Bk-1, Bk, and Bk+1 has the same configuration, the block Bk is used to describe the configuration of the string unit STU.

As shown in FIG. 31 and FIG. 32, in the present embodiment, the string unit STU is configured by arranging two NAND strings in plural in the direction of the word line. The two NAND strings are used as the source between the first and third selection gate lines D1 and D2 by sharing the first and third selection gate lines D1 and D2 as the drain side selection gate line SGD. The second selection gate line S2 of the gate selection gate line SGS, and the word line groups WLG1, WLG3 including the plurality of word lines WL arranged corresponding to the first and third selection gate lines D1 and D2, and The character line group WLG2 of the plurality of word lines WL arranged corresponding to the second selection gate line S2 is configured.

That is, the first and third selection gate lines D1, D2, the second selection gate line S2, and the first The character line groups WLG1 and WLG3 arranged corresponding to the third selection gate lines D1 and D2 and the word line group WLG2 arranged corresponding to the second selection gate line S2 are shared by the two U-shaped semiconductors SC, respectively. The two U-shaped semiconductors SC are composed of the first to fourth semiconductors CL1 to CL4 and the connection portion JP.

Next, the composition of the string unit STU will be specifically described. The word line groups WLG1, WLG2, and WLG3 are formed by stacking a plurality of word lines WL above the semiconductor substrate Ba. The first, second, and third character line groups WLG1, WLG2, and WLG3 and the first, second, and third selection gate lines D1, S2, and D2 are arranged in a direction orthogonal to the bit line BL.

The columnar first semiconductor CL1 is inserted through the first selection gate line D1 and the word line group WLG1. One end of the first semiconductor CL1 is connected to the bit line BL1.

The columnar second semiconductor CL2 is inserted through the second selection gate line S2 and the word line group WLG2. As shown in FIG. 33, the second semiconductor CL2 is disposed at the same position as the first semiconductor CL1 in the word line direction. One end of the second semiconductor CL2 is connected to the source line SL. The other end of the second semiconductor CL2 is electrically connected to the other end of the first semiconductor CL1 via a connection portion JP formed in the semiconductor substrate Ba.

The columnar third semiconductor CL3 is penetrated through the second selection gate line S2 and the word line group WLG2. As shown in FIG. 33, the third semiconductor CL3 is disposed at a position shifted from the second semiconductor CL2 in the direction of the word line. One end of the third semiconductor CL3 is connected to the source line SL.

The columnar fourth semiconductor CL4 is penetrated through the third selection gate line D2 and the word line group WLG3. As shown in FIG. 33, the fourth semiconductor CL4 is disposed at the same position as the third semiconductor CL3 in the word line direction. One end of the fourth semiconductor CL4 is connected to the bit line B2. The other end of the fourth semiconductor CL4 is electrically connected to the other end of the third semiconductor CL3 via a connection portion JP formed in the semiconductor substrate Ba.

Selective transistors are formed at intersections of the first, second, third, and fourth semiconductors CL1 to CL4 and the first, second, and third selection gate lines D1, S2, and D2, and are first and second. 2. The third and fourth semiconductors CL1 to CL4 and the word line groups WLG1, WLG2, and WLG3 The intersection point forms a memory cell.

Each string of cells constitutes a logical block and is managed by the first logical block address. Alternatively, the semi-logic block may constitute a string unit including one half of one source line, and the semi-logic block is defined as the second logical block address.

Further, the configuration of the logical block is not limited to the configuration of each of the above embodiments. The logical block can also be set in the manner shown in FIG. Fig. 34 shows a plurality of string units connected to one bit line (not shown). Therefore, a plurality of memory cells (not shown) sharing a respective word line are actually arranged in the direction of the vertical paper surface.

In FIG. 34, among a plurality of memory cells in which each string unit is connected in series, for example, six memory cells adjacent to each other constitute word line groups WLG1 to WLGp and WLGp+1 to WLG2p. The common word line groups WLG1 to WLGp and WLGp+1 to WLG2p of each string unit respectively constitute logical blocks. Therefore, in the case of this example, there are 2p logical blocks. Figure 34 is a representative representation of a logical block composed of a word line group WLG1.

<Effects of the seventh embodiment>

According to the seventh embodiment described above, when the memory device of the BiCS flash memory of the seventh embodiment is used, the same effects as those of the above embodiments can be obtained.

(Eighth embodiment)

Next, a nonvolatile semiconductor memory device according to the eighth embodiment will be described. In the eighth embodiment, a case where a film containing carbon as a main component is applied to a charge accumulation layer of a planar type so-called floating gate type NAND flash memory will be described. In the eighth embodiment, the components having the same functions and configurations as those of the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

<Overall structure of NAND flash memory>

The configuration of the NAND flash memory 300 of the eighth embodiment will be schematically described with reference to FIG. Fig. 35 is a block diagram showing the basic configuration of the NAND flash memory 300 of the eighth embodiment.

As shown in FIG. 35, the NAND flash memory 300 includes a memory cell array 71, a row decoder 72, a data input/output buffer 73, a data input/output terminal 74, a column decoder 75, a control circuit 76, and a control signal input. Terminal 77, source line control circuit 78, well control circuit 79, and plane switch 80.

The memory cell array 71 includes a plurality of bit lines BL, a plurality of word lines WL, and a source line SL. The memory cell array 71 is composed of a plurality of blocks BLK in which electrically memorable memory cell transistors (also simply referred to as memory cells or the like) are arranged in a matrix. The memory cell transistor MT has, for example, a gate electrode including a control gate electrode and a charge accumulation layer (for example, a floating gate electrode), and a threshold value of the transistor according to the amount of charge injected into the floating gate electrode Changes, memory binary or multi-value data. Further, the memory cell transistor MT may be a MONOS (Metal-Oxide-Nitride-Oxide-Silicon) structure having electrons trapped in a nitride film.

The row decoder 72 has a sense amplifier (not shown) that senses the voltage of the bit line BL in the memory cell array 71, and a data memory circuit (not shown) that latches data for writing. Wait. The row decoder 72 reads the data of the memory cell transistor MT in the memory cell array 71 via the bit line BL, or detects the state of the memory cell transistor MT via the bit line BL, or via the bit line BL. A write control voltage is applied to the memory cell transistor MT and the memory cell transistor MT is written.

Further, the row decoder 72 selects the data memory circuit in the row decoder 72, and reads the data of the memory cell transistor MT read in the data memory circuit from the data output terminal 74 to the outside via the data output buffer 74. (host) output.

The data output buffer 74 receives data from the data output terminal 74 and is stored in the data memory circuit selected by the row decoder 72. Further, the data input/output buffer 73 outputs data to the outside via the data output/output terminal 74.

The data input/output terminal 74 receives various commands and addresses such as writing, reading, erasing, and status reading in addition to writing data.

The column decoder 75 selects any of the blocks BLK in the data reading operation, the writing operation, or the erasing operation, and sets the remaining block BLK to be non-selected. That is, the column decoder 75 applies a voltage necessary for the read operation, the write operation, or the erase operation to the word line WL and the selection gate lines VSGS and VSGD of the memory cell array 71.

The source line control circuit 78 is connected to the memory cell array 71. The source line control circuit 78 controls the voltage of the source line SL.

The well control circuit 79 is connected to the memory cell array 71. The well control circuit 79 controls the voltage of the semiconductor substrate (well) that forms the memory cell transistor MT.

The control circuit 76 controls the memory cell array 71, the row decoder 72, the data input/output buffer 73, the column decoder 75, the source line control circuit 78, and the well control circuit 79. The control circuit 76 includes, for example, a voltage generating circuit 76-1 that boosts the power supply voltage. The control circuit 76 boosts the power supply voltage by the voltage generating circuit 76-1 as needed, and applies the boosted voltage to the row decoder 72, the data output/input buffer 73, the column decoder 75, and the source line control circuit. 78.

The control circuit 76 is based on a control signal (instruction latch enable signal CLE, address latch enable signal ALE, ready/busy signal RY/BY, etc.) input from the outside via the control signal input terminal 77, and a self-output terminal. The control operation is performed by a command input from the data output/input buffer 73. That is, the control circuit 76 generates a desired voltage during programming, verification, reading, and erasing of the data based on the control signal and the command, and supplies it to each unit of the memory cell array 71.

The planar switch 80 is connected to the control circuit 76 and the voltage generating circuit 76-1. The plane switch 80 switches the output target of the voltage from the voltage generating circuit 76-1 based on a signal from the control circuit 76 or the like.

In addition, the composition of the memory cell array 71 is disclosed in U.S. Patent Application Serial No. 12/397,711 (filed on March 3, 2009), U.S. Patent Application Serial No. 13/451,185 (filed on April 19, 2012), Patent Application No. 12/405,626 (applied on March 17, 2009), U.S. Patent Application Serial No. 09/956,986 (filed on September 21, 2001). This application contains the entire contents of these U.S. patent applications.

<Configuration of Driver of Eighth Embodiment>

Fig. 36 is a block diagram for schematically showing the relationship between the CG driver and the plane switch of the eighth embodiment.

In Fig. 36, for the sake of simplicity, the case where the memory cell array 71 has two planes will be described. Further, in the present embodiment, a case where one plane has four blocks will be described.

As shown in FIG. 36, the voltage generating circuit 76-1 includes a power supply 761, a CG driver 762, and an SG driver 763. The power supply 761 supplies power to the CG driver 762, the SG driver 763, and other circuits.

As shown in FIG. 37, the CG driver 762 of the sixth embodiment includes a VCGSEL circuit 762a, CGN drivers 762b and 762d (16 in total), a CGD driver 762c (two in total), and a CGU driver 762e.

The VCGSEL circuit 762a outputs a voltage VPGM or VCGRV as a voltage VCGSEL based on a control signal from the control circuit 76.

The CGN drivers 762b and 762d and the CGD driver 762c output voltages of any of the voltages VCGSEL, VUSEL1, VUSEL2, and VSS based on a control signal from the control circuit 76.

The CGU driver 762e outputs a voltage of any of the voltages VUSEL1, VUSEL2, and VSS in accordance with a control signal from the control circuit 76. The voltages VCELSRC and VCPWELL are connected to the memory cell array 71.

The CGN driver drives the word line WL (also referred to as DataWL) that stores data in units of one. Further, each of the CGN drivers includes a CGNA driver having four drivers, a CGNB driver, a CGNC driver, and a CGND driver. The CGD driver drives the word line WL (also referred to as DummyWL) that does not store data in units of one. DummyWL is for the lithography margin and to ensure the cell characteristics of DataWL and prepare 0 pieces for each generation as needed. on. The CGD driver is equipped with a CGDD driver and a CGDS driver. The CGDD selects an output such as a voltage VREADK. Also, the CGDS selects and outputs VREAD. The CGU driver is a driver that has a selectable voltage but a driving force. In the program/read operation associated with the memory, the word line WL farther from the selected word line WL may be driven at the same potential. In this case, a CGU driver is used.

Further, the SG driver 763 is a driver that supplies electric power to a selection gate or the like of the memory cell array 71.

The plane switch 80 is provided with a plane switch CCSSW and a plane switch SGSW on each plane of the memory cell array 71. More specifically, the plane switch 80 includes a plane switch CGSW 801a and a plane switch SGSW 801b corresponding to the plane <0>, and includes a plane switch CCSW 802a and a plane switch SGSW 802b corresponding to the plane <1>.

The plane switch CCSW801a receives the zone signals ZONE_P0<3:0>, mode signals MODE_P0<1:0>, and CGD*SW_P0 from the control circuit 76. Further, the plane switch CCSW801a receives CGNA<3:0>, CGNB<3:0>, CGNC<3:0>, CGND<3:0>, CGDD, CGDS, and CGU from the CG driver 762. Further, the plane switch CCSW 801a supplies a signal from the CG driver 762 to the column decoder 75 based on a signal from the control circuit 76. Further, the plane switch SGSW 801b supplies the SGS signal and the SGD signal received from the SG driver 763 to the column decoder 75 based on the signal from the control circuit 76.

The column decoder 75 is provided with a column decoder for each plane. More specifically, the column decoder 75 includes a column decoder 751 corresponding to the plane <0> and a column decoder 752 corresponding to the plane <1>.

Column decoder 751 receives signals BLKADD_P0<1:0> and signal RDEC_P0 from control circuit 76. Further, the column decoder 751 receives CGI<31:0>, CGDDI, and CGDSI from the plane switch CGSW801a. Furthermore, column decoder 751 receives SGSI, SGDI, USGSI, and USGDI from plane switch SGWW 801b. Column decoder 751 is based on The signal is received and the signal is supplied to plane <0>. Further, the column decoder 752 operates in the same manner as the column decoder 751.

<Configuration of CGD-related switches of the planar switch CGSW>

Next, the configuration of the CGD-related switch of the planar switch CCSW of the eighth embodiment will be schematically explained using FIG.

For example, in the eighth embodiment, the plane switch CCSW801a includes switches 80a, 80b, 80c, and 80d.

The CGDD is input to one end of the voltage path of the switch 80a, the other end of the voltage path is connected to the signal line CGDDI, and the signal from the control circuit 76 is input to the gate.

The CGDS is input to one of the voltage paths of the switch 80b, the other end of the voltage path is connected to the signal line CGDDI, and the signal from the control circuit 76 is input to the gate.

The CGDD is input to one of the voltage paths of the switch 80c, the other end of the voltage path is connected to the signal line CGDSI, and the signal from the control circuit 76 is input to the gate.

The CGDS is input to one of the voltage paths of the switch 80d, the other end of the voltage path is connected to the signal line CGDSI, and the signal from the control circuit 76 is input to the gate.

In addition, the configuration of the switch related to the CGN of the plane switch CGSW of the present embodiment is the same as the configuration of the switch related to the CGN of the plane switch CCSSW described in the first embodiment, and thus the description thereof is omitted.

<Example of CG mapping>

The CG map of the eighth embodiment will be schematically described using Figs. 39 to 41. Fig. 39 is a view showing a CG map at the time of a programming operation of the semiconductor memory device of the eighth embodiment. Fig. 40 is a view showing a CG map at the time of a reading operation of the semiconductor memory device of the eighth embodiment. Fig. 41 is a view showing a CG map at the time of erasing operation of the semiconductor memory device of the eighth embodiment. In FIGS. 39 to 41, the vertical axis represents the allocation of the CG driver to the word line WL, and the horizontal axis represents the selected word line WL.

Further, in the eighth embodiment, the CGDD driver is always in addition to the reading operation. The word line WLDD applies a dedicated voltage, and the CGDS driver always applies a dedicated voltage to the word line WLDS.

<Example of CG mapping when programming action>

First, the CG mapping at the time of programming operation will be explained. As shown in FIG. 39, a CG driver that applies a voltage to the word line WL is appropriately switched in accordance with the selected word line WL.

The area shown on the horizontal axis of Fig. 39 is information from the control circuit indicating that each DataWL is connected to any of the CGN drivers or the CGU driver. For example, the plane and page address for accessing the memory controller 20 are input from the host 2. Therefore, by inputting the plane and page address for accessing the NAND flash memory 300 from the memory controller 20, the memory device internal control circuit 76 transmits ZONE by the plane switching circuit 80 of the plane. 3:0> and decided. Specifically, when the word lines WLDS, WL0 WL WL9 are programmed, the desired voltage is applied from the CGNA driver to the word lines WL0 WL3, and the desired voltage is applied from the CGNB driver to the word lines WL4 -7, respectively. The word lines WL8~11 apply a desired voltage from the CGNC driver, apply a desired voltage to the word lines WL12~15 from the CGND driver, and apply a desired voltage to the word lines WL16~31 from the CGU driver.

On the other hand, when the word lines WL10 to WL13 are programmed, the CGNA driver is connected to the word lines WL16 to 19, and the CGU is connected to the word lines WL0 to WL3. A predetermined connection is made to the word line WL to be programmed, and the combination of the connections is five types of the areas PZ0 to PZ4.

As described in the first embodiment, each of the areas PZ0 to PZ4 is referred to as a zone at the time of programming.

In the present embodiment, by using a total of 16 CGN drivers, the non-selected word line WL(i+1) can be controlled with high precision by the CGN driver by selecting the word line WLi (i: 0 to 31) during programming. ~ Non-selected word line WL(i+6) (refer to D6 in the figure) or unselected word line WL(i-1) to non-selected word line WL(i-6) (refer to S6 in the figure) Voltage.

<Example of CG mapping when reading action>

Next, the CG mapping at the time of the reading operation will be described.

When reading the NAND type semiconductor memory device, as long as a read voltage is applied to the selected word line WLi, a voltage VREADK is applied to the word line WL of the unselected word line WL (i±1), and the other word lines WL are applied. By applying a voltage called voltage VREAD, the range of word lines WL that must be controlled is narrower than that during programming, thereby reducing the number of necessary CGN drivers.

As shown in FIG. 40, a CG driver that applies a voltage to each word line WL is appropriately switched in accordance with the selected word line WL.

As shown on the horizontal axis of Fig. 40, the areas RZ0 to RZ6 at the time of reading are set.

Specifically, when reading the word lines WLDS, WL0 WL WL5, the desired voltage is applied to the word lines WL0 WL3 by the CGNA driver or the CGNC driver, and the CGNB driver or CGND is also applied to the word lines WL4 -7. The driver applies the desired voltage and applies the desired voltage to the word lines WL8-31 in a CGU driver. On the other hand, when the word lines WL6 to WL9 are read, switching is performed such that the CGNA driver or the CGNC driver is connected to the word lines WL8 to 11 and the CGU driver is connected to the word lines WL0 to WL3. A predetermined connection is made to the read word line WL, and the combination of the connections is seven types of the regions RZ0 to RZ6. Each of the regions RZ0 to RZ6 is referred to as a region at the time of reading.

When the zone RZ0 is selected, the zone signal becomes "000", and when the zone RZ1 is selected, the zone signal becomes "001". When the area RZ2 is selected, the area signal becomes "010", and when the area RZ3 is selected, the area signal becomes "011". When the area RZ4 is selected, the area signal becomes "100", and when the area RZ5 is selected, the area signal becomes "101". Also, in the case of selecting the zone RZ6, the zone signal becomes "110".

As described above, in the present embodiment, at least the non-selected word line WL(i+1) (refer to D1 in the drawing) or the non-selected word can be switched by the CGN driver with respect to the selected word line WLi (i: 0 to 31). The voltage of the line WL(i-1) (refer to S1 in the figure), and can be read by multi-plane by assigning the word line WL for different planes with CGNA and CGNB, CGNC and CGND The time is free to specify two word lines WL. Since the NAND type semiconductor memory device which can have, for example, 16 CGN drivers selects two word lines WL during multi-plane reading, the driver is divided into four groups, and two groups are allocated for selecting one word line WL, and the remaining The 2 group allocation is used to select another word line WL.

<Example of CG mapping when erasing action>

Next, the CG mapping at the time of erasing the action will be described.

As shown in FIG. 41, during the erase operation, the CGNA driver applies a voltage to the word lines WL0 to WL3, WL16 to WL19, and the CGNB driver applies a voltage to the word lines WL4 to WL7 and WL20 to WL23. Further, the CGNC driver applies a voltage to the word lines WL8 to WL11 and WL24 to WL27, and the CGND driver applies a voltage to the word lines WL12 to WL15 and WL28 to WL31. In addition, since this embodiment is not related to the erasing operation, detailed description is omitted.

<CG connection form>

Next, a connection table of CG will be described using FIG. 42A and FIG. 42B. Fig. 42A shows the connection relationship from the CGN/CGU driver to the CGI for the erase signal, the program operation, and the read operation. Figure 42B shows the connection relationship from the CGD driver to CGD*I.

As shown in FIG. 42A, when erasing, the mode signal MODE<1:0> becomes "00", and when programming, the mode signal MODE<1:0> becomes "01". At the time of reading (Read-A), the mode signal MODE<1:0> becomes "10", and when reading (Read-B), the mode signal MODE<1:0> becomes "11". In the reading (Read-A) and reading (Read-B), although the reading operation itself is substantially unchanged, the CG drivers used are different.

As shown in Fig. 42B, when CGDDSW is "0", the CGDDI output is the output of the CGDD driver, and when CGDDSW is "1", the CGDDI output is the output of the CGDS driver. Further, when CGDDSSW is "0", the CGDSI output is the output of the CGDS driver, and when the CGDDSSW is "1", the CGDSI output is the output of the CGDD driver.

<Effects of the eighth embodiment>

According to the eighth embodiment described above, in the memory device using the planar NAND flash memory, the same effects as those of the above embodiments can be obtained.

(Ninth Embodiment)

Next, a ninth embodiment will be described. In the ninth embodiment, the reading operation in the vicinity of the dummy word line of the NAND flash memory 300 described in the eighth embodiment will be described. In the ninth embodiment, the components having the same functions and configurations as those of the above-described eighth embodiment are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

The reading operation when the word line WL is selected in the vicinity of the dummy word line WLD and the reading operation in the case where the selected word line WL is not in the vicinity of the dummy word line WLD will be described with reference to FIGS. 43A and 43B. 43A is a view showing a region signal ZONE<3:0>, a mode signal MODE<1:0>, a CGDDSW signal, and a CGDDSSW signal in the case where the word line WL is selected in the vicinity of the dummy word line WLD in the read operation. . Fig. 43B is a view showing the kind of the CG driver used for each word line WL and the voltage applied to the word line WL. In addition, here, for the sake of simplicity, attention is paid to the case where the plane 0 and the plane 1 are read, and the case where the hSLC data is read from the plane 1 will be described.

As shown in FIG. 43A and FIG. 43B, in the case of the plane 0, when the word line WL is selected, for example, in the case of the word line WL31 adjacent to the dummy word line WLDD, the area signal ZONE<3:0> is "110". The mode signal MODE<1:0> is "10", the CGDDSW signal is "0", and the CGDSSW signal is "0".

For the selected word line WL31, the voltage VCGRV is applied by the CGNB<3> driver, and the voltage VREADK (VREADK>VCGRV) is applied to the unselected word line WL30 by the CGNB<2> driver. Next, for the dummy select word line WLDD, a voltage VREADK (VREADK > VCGRV) is applied by the CGDD driver.

As shown in FIG. 43A and FIG. 43B, in the plane 1, the selected word line WL is, for example, not adjacent. When the word line WL15 of the dummy word line WLD is connected, the area signal ZONE<3:0> is "011", the mode signal MODE<1:0> is "11", the CGDDSW signal is "1", CGDSWS The signal is "0".

For the selected word line WL15, the voltage VCGRV is applied by the CGND<3> driver, and the voltage VREADK (VREADK>VCGRV) is applied to the non-selected word lines WL14 and WL16 by the CGND driver.

(Tenth embodiment)

Next, a tenth embodiment will be described. In the tenth embodiment, a CG driver and a power supply device different from the CG driver and the power supply device described in the ninth embodiment will be described. In the tenth embodiment, the components having the same functions and configurations as those of the above-described embodiments are denoted by the same reference numerals, and the description thereof will be repeated only when necessary.

As shown in Fig. 44, the power supply 761 and the CG driver 762 of the tenth embodiment divide the power supply into a plane A and a plane B. As shown in FIG. 44, the CG driver 762 of the tenth embodiment includes a VCGSEL circuit 762a, CGN drivers 762b and 762d (16 in total), a CGD driver 762c (two in total), a CGU driver 762e, a VCGSEL2 circuit 762f, and a CGD driver 762g. (2 in total), and CGU driver 762h.

The VCGSEL circuit 762a outputs a voltage VPGM or VCGRVA as a voltage VCGSEL_AB in accordance with a control signal from the control circuit 15.

The CGN driver 762b and the CGD driver 762c output any of the voltages VCGSEL_AB, VUSEL1A, VUSEL2A, and VSS to the plane A based on the control signal from the control circuit 76.

The CGU driver 762e outputs any of the voltages VUSEL1A, VUSEL2A, and VSS to the plane A based on the control signal from the control circuit 76.

The VCGSEL2 circuit 762f outputs voltages VPGM, VCGRVA, and VCGRVB as voltages VCGSEL_CD based on control signals from the control circuit 76. also, The VCGSEL2 circuit 762f outputs voltages VUSEL1A and VUSEL1B as voltages VUSEL1_CD in accordance with a control signal from the control circuit 76. Further, the VCGSEL2 circuit 762f outputs the voltages VUSEL2A and VUSEL2B as the voltage VUSEL2_CD based on the control signal from the control circuit 76.

The CGN driver 762d and the CGD driver 762g output any voltages of the voltages VCGSEL_CD, VUSEL1_CD, VUSEL2_CD, and VSS to the plane B based on the control signal from the control circuit 76.

The CGU driver 762h outputs any of the voltages VUSEL1_CD, VUSEL2_CD, and VSS to the plane B based on the control signal from the control circuit 76. The voltages VCELSRCA and VCPWELLA are connected to the memory cell array 71 of the plane A. The voltages VCELSRCB and VCPWELLB are connected to the memory cell array 71 of the plane B. In addition, the plane A and the plane B may be any plane.

<Effects of the tenth embodiment>

According to the tenth embodiment, the power supply 761 of the tenth embodiment has two voltage systems for two planes as compared with the power supply 761 of the ninth embodiment, and that voltage can be applied to both planes simultaneously. The CG driver is constructed. Therefore, for example, the above-mentioned MLC data, and hSLC data, or SLC and hSLC data can be simultaneously read.

(changes, etc.)

Further, the first to eighth embodiments can be combined in various combinations.

Further, in each of the above embodiments, the CGN driver is provided with 16 CGNA<3:0>, CGNB<3:0>, CGNC<3:0>, and CGNC<3:0>, but is not necessarily limited. herein. As described in each of the above embodiments, a CGN driver that can adjust the voltage applied to the unselected word line WL near the selected word line WL may be used. Further, in each of the above embodiments, the CG mapping is displayed, but it is only an example. According to the gist of the above embodiments, the number of CGN drivers may be increased or decreased. Change the CG mapping as appropriate.

Further, in the second and third embodiments described above, the BiCS flash memory is used for the description, but the same effect can be obtained in the case of the planar NAND flash memory.

Further, in each of the above embodiments, a memory cell of two or four values is described, but the present invention is not limited thereto and can be appropriately changed.

Further, the range of the area described in each of the above embodiments is merely an example, and the range of the area can be appropriately changed.

Further, in the above embodiments, the programming method using the hSLC mode or the like is described, but the present invention is not limited thereto, and the SLC mode may be used instead of the hSLC mode.

Further, in each of the above embodiments, the memory cell array 11 has a plane <0> and a plane <1>. However, the memory cell array 11 is not limited to this, and the memory cell array 11 can also hold a specific number of planes.

The embodiments of the present invention have been described above, but the present invention is not limited to the embodiments described above, and various modifications can be made without departing from the spirit and scope of the invention. Furthermore, the above embodiments include various stages of the invention, and various inventions can be obtained by appropriately combining the disclosed constituent elements. For example, if a plurality of constituent elements are removed from the disclosed constituent elements and a specific effect is still obtained, the invention can also be taken as an invention.

11‧‧‧ Memory Cell Array

15‧‧‧Control circuit

16‧‧‧Voltage generation circuit

17‧‧‧ plane switch

19‧‧‧Output buffer

21‧‧‧ column decoder

161‧‧‧Power supply

162‧‧‧CG driver

163‧‧‧SG driver

171a‧‧‧Plane switch

171b‧‧‧ plane switch

172a‧‧‧Plane switch

172b‧‧‧ plane switch

211‧‧‧ column decoder

212‧‧‧ column decoder

BG‧‧‧ back gate

BLK‧‧‧ Block

CGBG‧‧‧ drive

CGBGI‧‧‧ signal line

CGDDB‧‧‧ drive

CGDDBI‧‧‧ signal line

CGDDT‧‧‧ drive

CGDDTI‧‧‧ signal line

CGDSB‧‧‧ drive

CGDSBI‧‧‧ signal line

CGDST‧‧‧ drive

CGDSTI‧‧‧ signal line

CGI‧‧‧ signal line

CGNA‧‧‧ drive

CGNB‧‧‧ drive

CGNC‧‧‧ drive

CGND‧‧‧ drive

CGU‧‧‧ drive

SGD‧‧‧汲polar selection gate line

SGDI‧‧‧ signal line

SGS‧‧‧Source side selection gate line

SGSI‧‧‧ signal line

USGDI‧‧‧ signal line

USGSI‧‧‧ signal line

VSS‧‧‧ voltage

WL‧‧‧ character line

WLD‧‧‧Dummy word line

Claims (19)

A memory system comprising: a memory device, comprising: a plurality of memory cells each holding data; a plurality of word lines connected to the plurality of memory cells; and a page having a connection to the same a plurality of memory cells of the word line; a plane having the plurality of pages; a memory cell array having a plurality of the planes; and a plurality of word line drivers for applying voltage to the plurality of word lines And a controller that controls the memory device; and the controller issues an instruction to the memory device, the command simultaneously performing reading of the first data on the specific plane and reading from a plane different from the specific plane The second information. The memory system of claim 1, wherein the memory device receives the first data read command and the second data read command from the controller, from the specific plane and a plane different from the specific plane Read the data at the same time. The memory system of claim 1, wherein the memory device further comprises a plurality of switches disposed on each of the planes and allocating the word line driver to each of the word lines. The memory system of claim 3, wherein said controller integrates a plurality of accesses to said different planes different from each other when said plurality of access requests are made to said memory cell array, and said word line driver is used And the switch, wherein the plurality of word line drivers are allocated to the plurality of word lines on each of the planes. The memory system of claim 3, wherein the controller, in the case of having a plurality of access requests to the memory cell array, integrates pages belonging to the above-mentioned planes different from each other and belonging to the above-mentioned word lines different from each other And a plurality of accesses, wherein the plurality of word line drivers are allocated to the plurality of word lines on each of the planes using the word line driver and the switch. The memory system of claim 3, wherein the controller integrates a plurality of accesses to different planes and associated data, and uses the above characters when there are multiple access requests to the memory cell array. The line driver and the switch respectively distribute the plurality of word line drivers to the plurality of word lines on each of the planes. The memory system of any one of claims 4 to 6, wherein the controller integrates the plurality of accesses associated with reading of the data. The memory system of any one of claims 4 to 6, wherein the controller integrates the plurality of accesses associated with programming of the data. The memory system of claim 1, further comprising a power supply for supplying voltage to the two of said planes; and said plurality of word line drivers simultaneously supplying the two said planes with a voltage supplied from said power supply. The memory system of claim 1, wherein the controller stores the first data and the second data on mutually different planes when it is determined that the first data supplied from the outside is associated with the second data. The memory system of claim 1, wherein in the plurality of planes, the MLC data is stored in the first plane, and the hSLC data having a higher threshold than the SLC data is stored in the second plane, the foregoing control At the same time, the MLC data stored in the first plane described above is read and stored. hSLC data in the second plane above. The memory system of claim 1, wherein in the plurality of planes, the SLC data is stored in the first plane, and the hSLC data having a higher threshold than the SLC data is stored in the second plane, the foregoing control The device simultaneously reads the SLC data stored in the first plane and the hSLC data stored in the second plane. The memory system of claim 1, wherein in the plurality of planes, when the MLC data is stored in the first plane and the SLC data is stored in the second plane, the controller is simultaneously read and stored in the first The MLC data of the plane and the SLC data stored in the second plane above. The memory system of claim 1, wherein the controller further comprises a queue region that receives the plurality of access requests. The memory system of claim 1, wherein the controller reads a plurality of data from a plurality of word lines belonging to the plurality of planes in a data reading order. The memory system of claim 3, wherein the controller further comprises a memory portion that retains range information of the word line; and controls the switch based on the range information. The memory system of claim 1, wherein the memory cell array is a 3-dimensional laminated non-volatile semiconductor. The memory system of claim 1, wherein the memory cell array is a NAND flash memory. A memory system comprising: a memory device, comprising: a plurality of memory cells each holding data; a plurality of word lines connected to the plurality of memory cells; and a page having a connection to the same a plurality of memory cells of the word line; a plane having the plurality of pages; a memory cell array having a plurality of the planes; and a plurality of word line drivers for applying voltage to the plurality of word lines; and a controller for controlling the memory device And the above controller simultaneously accesses the pages different from each other described above.