CN110164487A - A kind of framework of dynamic RAM - Google Patents

A kind of framework of dynamic RAM Download PDF

Info

Publication number
CN110164487A
CN110164487A CN201910275295.6A CN201910275295A CN110164487A CN 110164487 A CN110164487 A CN 110164487A CN 201910275295 A CN201910275295 A CN 201910275295A CN 110164487 A CN110164487 A CN 110164487A
Authority
CN
China
Prior art keywords
block
memory
sub
reading
array
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201910275295.6A
Other languages
Chinese (zh)
Inventor
吴君
张学渊
朱光伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou Huifeng Microelectronics Co Ltd
Original Assignee
Suzhou Huifeng Microelectronics Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Suzhou Huifeng Microelectronics Co Ltd filed Critical Suzhou Huifeng Microelectronics Co Ltd
Priority to CN201910275295.6A priority Critical patent/CN110164487A/en
Publication of CN110164487A publication Critical patent/CN110164487A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/21Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
    • G11C11/34Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices
    • G11C11/40Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors
    • G11C11/401Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming cells needing refreshing or charge regeneration, i.e. dynamic cells
    • G11C11/4063Auxiliary circuits, e.g. for addressing, decoding, driving, writing, sensing or timing
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C5/00Details of stores covered by group G11C11/00
    • G11C5/02Disposition of storage elements, e.g. in the form of a matrix array
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C5/00Details of stores covered by group G11C11/00
    • G11C5/06Arrangements for interconnecting storage elements electrically, e.g. by wiring

Abstract

The invention discloses a kind of framework of memory, specifically a kind of low-power consumption double data rate dynamic random access memory (LPDDR).Belong to reservoir designs technical field.Purpose is to provide a kind of framework of low-power consumption double data rate dynamic random access memory (LPDDR).The framework has 8 or more memory block, command block, input/output block, data paths etc..The first sub-block and the second sub-block of each memory block are located in NxN array, and wherein N is an even number, wherein the first sub-block is located at the first half of NxN array, wherein the second sub-block is located at the latter half of NxN array.The first sub-block and the second sub-block of each memory block are located at the symmetrical position of array center position, so that the distance of each memory block of input and output block access is essentially identical, the power consumption of the reading and writing data order of final difference memory block lower, more preferable can be predicted, is more symmetrical.

Description

A kind of framework of dynamic RAM
Technical field
The invention discloses a kind of framework of memory, specifically a kind of low-power consumption double data rate dynamic random is deposited Access to memory (LPDDR).Belong to reservoir designs technical field.
Background technique
Synchronous Dynamic Random Access Memory (SDRAM) is the dynamic random access memory synchronous with computer system bus Device (DRAM).The especially sync cap of SDRAM, can waiting system clock signal before responding input control.Clock signal For driving internal finite state machine to carry out pipeline to incoming order.Data storage areas is divided into several pieces, Allow the multiple storage access commands of memory single treatment.This allows higher data access speed than asynchronous DRAM memory chip Rate.
Low-power consumption double data rate (DDR) SDRAM is one kind of DDR SDRAM, also known as mDDR (also referred to as low-power consumption DDR or LPDDR), it is the communication mark that the American Association EEE electronic equipment engineering committee (JEDEC) formulates towards low-power consumption RAM Standard, it is famous with low-power consumption and small size, dedicated for intelligent movable mobile phone, tablet computer and other mobile computing devices.
One typical LPDDR includes multiple memory blocks, input/output block, command block.Input/output block can be from specific Memory block in read data, can also input data into specific memory block.With the increase of memory capacity, each memory block Distance to input/output block increases, and the power consumption of reading and writing data can be increasing.
Summary of the invention
The object of the present invention is to provide a kind of framves of low-power consumption double data rate dynamic random access memory (LPDDR) Structure.With 8 or more memory blocks, command block, input/output block, data path etc..First sub-block of each memory block and Second sub-block is located in NxN array, and wherein N is an even number, wherein the first sub-block is located at the first half of NxN array, wherein Second sub-block is located at the latter half of NxN array.The first sub-block and the second sub-block of each memory block are located at array center position Symmetrical position so that the distance of each memory block of input and output block access is essentially identical, final difference memory blocks The power consumption of reading and writing data order lower, more preferable can be predicted, is more symmetrical.
Low-power consumption dynamic RAM of the invention, feature include a command block, the memory with memory block Core, at least two reading and writing data channels, one outputs and inputs block:
Each memory block at least has the first sub-block and the second sub-block;Wherein the first sub-block memory block is connected to reading and writing data channel In first data read/write channel, it is logical that the second sub-block memory block is connected to second reading and writing data in reading and writing data channel Road;
Reading and writing data channel is in memory core and outputs and inputs and transmits data between block, wherein command block drive command signal It operates memory core, reading and writing data channel and outputs and inputs block;
First the second sub-block of sub-block is located in NxN array, and wherein N is an even number, wherein the first sub-block is located at the first half of array Part, wherein the second sub-block is located at the latter half of array;
Command block is arranged along the 1st side of array, is output and input block and is arranged along the 2nd side of array, wherein the 1st side and the 2nd side position In the relative position of array;
First reading and writing data channel is routed between the first half of array, latter half of the second reading and writing data channel in array Between be routed;
The first sub-block and the second sub-block of each memory block are located in array, positioned at the symmetrical position of array center position It sets;
For the data that memory is read, data word has the first range byte, the second range byte, third range byte, 4th range byte, wherein first outputs and inputs block the first range byte of processing, second, which outputs and inputs block, handles the second model Byte is enclosed, third outputs and inputs block processing third range byte, and the 4th, which outputs and inputs block, handles the 4th range byte.Wherein First reading and writing data channel, which is connected to first and outputs and inputs block and second, outputs and inputs block, the connection of the second reading and writing data channel Block and the 4th, which is output and input, to third outputs and inputs block.
For the data for being written to memory, the data of memory are read from block input is output and input by data Write access is to some memory block.Data can be divided into first group and second group of data, wherein first group of data is stored in some First sub-block of particular memory block, second group of data are stored in the second sub-block of some particular memory block.
Specific structure of the invention is disclosed in detail below.
A kind of framework of dynamic RAM, including a command block, the memory with memory block, it is characterized in that often A memory block at least has the first sub-block and the second sub-block;
It further include the channel at least two reading and writing datas (DRW), wherein the first sub-block memory block is connected at least two reading and writing datas First data read/write channel in channel, and the second sub-block memory block therein are connected at least two reading and writing data channels Second data read/write channel;
And one output and input (DQ) block, wherein at least two reading and writing data channel passes between memory core and DQ block Transmission of data, wherein command block drive command signal operation memory core, reading and writing data channel and DQ block.
In the memory block of the memory, first the second sub-block of sub-block is located in NxN array, and wherein N is an idol Number, wherein the first sub-block is located at the first half of array, wherein the second sub-block is located at the latter half of array.
The command block arranges that wherein DQ block is arranged along the 2nd side of array along the 1st side of array, wherein the 1st side and the 2 sides are located at the opposite of array.
The first reading and writing data channel is routed between the first half of array, and the second reading and writing data channel is in array Latter half between be routed.
The first sub-block and the second sub-block of the memory block are located in array, the bilateral symmetry positioned at array center position Position.
The DQ block is arranged along the 2nd side of array.
The wherein data word has the first range byte, the second range byte, third range byte, the 4th range word Section, wherein the first DQ block handles the first range byte, the 2nd DQ block handles the second range byte, and the 3rd DQ block handles third range Byte, the 4th DQ block handle the 4th range byte.Wherein the first reading and writing data channel is connected to the first DQ block and the 2nd DQ block, the Two reading and writing data channels are connected to the 3rd DQ block and the 4th DQ block.
Data are inputted from DQ block, pass through reading and writing data channel to some memory block.Data can be divided into first group and Two groups of data, wherein first group of data is stored in the first sub-block of some particular memory block, second group of data is stored in some spy Determine the second sub-block of memory block.
The first sub-block and the second sub-block of each memory block of the invention are located at the symmetrical position of array center position It sets, so that the distance of each memory block of input and output block access is essentially identical, the reading and writing data order of final difference memory block Power consumption lower, more preferable can be predicted, is more symmetrical.
Detailed description of the invention
Attached drawing 1 is the memory block diagram that the present invention has storage sub-block.
Attached drawing 2 is the block diagram of the one embodiment for the memory that the present invention has more memory blocks.
Attached drawing 3 is memory one embodiment block diagram with more storage sub-blocks of the invention.
Attached drawing 4 is the additional embodiment block diagram of the memory with more memory blocks of the invention.
Attached drawing 5 is another embodiment of the memory with storage sub-block of the invention.
Attached drawing 6 is the data path of memory of the invention.
Specific embodiment
Embodiment one
In the detailed description of following embodiment, the attached drawing in this document is referred to, these attached drawings can specifically be practiced Embodiment.
A kind of framework of dynamic RAM, such as DRAM, including the region with multiple memory blocks.These memory blocks Region is properly termed as storage core.Usual data read/write bus can be individual data reading (DR) line and data write (DW) line, Or multiplexed data reads and writes (DRW) line.
A kind of framework of dynamic RAM, including a command block, the memory with memory block, it is characterized in that often A memory block at least has the first sub-block and the second sub-block;
It further include the channel at least two reading and writing datas (DRW), wherein the first sub-block memory block is connected at least two reading and writing datas First data read/write channel in channel, and the second sub-block memory block therein are connected at least two reading and writing data channels Second data read/write channel;
And one output and input (DQ) block, wherein at least two reading and writing data channel passes between memory core and DQ block Transmission of data, wherein command block drive command signal operation memory core, reading and writing data channel and DQ block.
In the memory block of the memory, first the second sub-block of sub-block is located in NxN array, and wherein N is an idol Number, wherein the first sub-block is located at the first half of array, wherein the second sub-block is located at the latter half of array.
The command block arranges that wherein DQ block is arranged along the 2nd side of array along the 1st side of array, wherein the 1st side and the 2 sides are located at the opposite of array.
The first reading and writing data channel is routed between the first half of array, and the second reading and writing data channel is in array Latter half between be routed.
The first sub-block and the second sub-block of the memory block are located in array, the bilateral symmetry positioned at array center position Position.
The DQ block is arranged along the 2nd side of array.
The wherein data word has the first range byte, the second range byte, third range byte, the 4th range word Section, wherein the first DQ block handles the first range byte, the 2nd DQ block handles the second range byte, and the 3rd DQ block handles third range Byte, the 4th DQ block handle the 4th range byte.Wherein the first reading and writing data channel is connected to the first DQ block and the 2nd DQ block, the Two reading and writing data channels are connected to the 3rd DQ block and the 4th DQ block.
Data are inputted from DQ block, pass through reading and writing data channel to some memory block.Data can be divided into first group and Two groups of data, wherein first group of data is stored in the first sub-block of some particular memory block, second group of data is stored in some spy Determine the second sub-block of memory block.
One typical DRAM may include eight or more memory block, such as block 0, block 1 etc..Memory core is by 8 A or 8 or more memory blocks form.Each piece may include bit line, wordline, storage unit, bit line detection amplifier, part With global row decoder, column decoder etc..Each memory block can respond the data read/write order to the memory block.
Fig. 1 is the memory block diagram that the present invention has storage sub-block.Memory of the invention includes a command block 10, interior Deposit core 12, reading and writing data channel 22 and 24 and input/output (DQ) 14.Reading and writing data channel 22 and 24 is multiplexing Reading and writing data bus.
Order and address block 10 can also include test logic 16 and 20, order and address logic 18.Test 16 He of logic 20 include the internal test mode control circuit for testing respective memory.Order and address logic 18 are included as respective deposit The circuit of reservoir processing order and address.
Storing core 12 includes eight block storages: memory block 0,1,2,3,4,5 67.Each memory block of 0-7 can wrap Include first sub-block and second sub-block.The first sub-block and the second sub-block of 0-7 memory block are logical by reading and writing data respectively Road 22 and 24 is connected with DQ block 14.Half of sub-block of 0-7 memory block is arranged by 4x4 array.First sub-block cloth of 0-7 memory block Set the top half in 4x4 array, i.e. the 1st row and the 2nd row of array.Second sub-block of 0-7 memory block is arranged in 4x4 array Lower half portion, i.e. the 3rd row and the 4th row of array.Reading and writing data channel 22 is routed between the 1st row of 4x4 array and the 2nd row, The first sub-block of memory block 0-7 is connected to DQ block 14.The cloth between the 3rd row of 4x4 array and the 4th row of reading and writing data channel 24 Line is connected to the second sub-block of memory block 0-7 to DQ block 14.
If Subset block array is extended to NxN Subset block array, every two rows sub-block may be coupled to different data and read Write access.Each reading and writing data channel can be located between every two row of NxN array, therefore upper sub-block and lower sub-block are to respective The distance in reading and writing data channel is essentially identical.By keeping the distance for accessing each memory block essentially identical, in different memory blocks The power consumption of reading and writing data order can be predicted more preferably and more symmetrical.
The preceding half block and rear half block of 0-7 memory block are arranged separately in the array relative to DQ block 14, deposit access each Store up difference of the relative distance between different memory blocks of block and little.For the Subset block array of a 4x4, if some is stored One sub-block of block is located at the first row of array, then second sub-block of the memory block will be located at the 4th column.If some is deposited The sub-block of storage block is located at the secondary series of array, arranges then other sub-blocks of the memory block will be located at third.2 of each memory block Sub-block is located at the central symmetry position of array.
Such as the first sub-block of memory block 0 is located at the first row of array, the second sub-block of memory block 0 is located at the 4th of array the In column (or last column).When data are read in memory block 0, while two sub-blocks of access block 0.Data can be passed through Read/write channel 22 reads the partial data of the first sub-block, and the partial data of the second sub-block is read by data read/write channel 24. Since there are sub-block, each sub-block to respective reading and writing data channel 22 and 24 in the two sides of the physical location of reading and writing data channel 22 and 24 Distance it is all identical.To realize, whichsoever memory block is accessed, and power consumption is all similar.Because storing in core 12 The total distance of each memory block to DQ block 14 is similar.
Under the extreme case of not above scheme, if the first sub-block and the second sub-block of memory block 0 all in column 1, Compared with other memory blocks not in first row, data must could be incited somebody to action by distance farthest on reading and writing data channel 22 and 24 Data are passed to/taking-up memory block 0.If the first sub-block and the second sub-block of memory block 0 are all in column 4, relative to other not Data must be passed to/taking-up memory block 0 by shortest distance on reading and writing data channel by the memory block 7 of four column, these sub-blocks. Due to these different distances, so that memory block 0 is with memory block 7 because data are passed to/take out the transmission power consumption generated and have very Big difference causes power consumption asymmetric.
Storing core 12 can further comprise simulated block 40, read with providing various reference voltages to block storage 0-7 and data/ Write access 22 and 24.Analog circuit block 40 may include the typical element of memory, including voltage generator, detector, other electricity Source generating circuit, clamp circuit etc..According to the design and specification of memory, simulated block 40 can provide other power consumptions of memory It is required that.
The people with common skill is appreciated that the sub-block of any amount of memory block and/or any scale is equal in this respect It is applicable to the present invention.In order to help to understand that this is invented, eight memory blocks are each there are two half block, are described to convey this The concept of invention.However, the memory block of other quantity and/or any amount of sub-block are used equally for the present invention.Such as it can make With 16 memory blocks, each memory block has 4 sub-blocks (i.e. 1/4 sub-block).Other variations of memory block and/or sub-block quantity It is included in the present invention.In addition, the array size of sub-block equally can extend to NxN array, wherein N can be even number.Number It can be routed between every two rows sub-block according to read/write channel and be connected to DQ.
Data are transmitted between DQ block 14 and memory block 0-7 in reading and writing data channel 22 and 24.DQ block 14 includes DQ control block 32 and DQ block 28,30,34 and 36.The data that DQ block 28,30,34 and 36 separately includes specified byte unit output and input electricity Road.When DQ control block 32 can control 28 to DQ block, 30,34 and 36 global write (data write-in) and reading (data output) Sequence.Therefore, DQ block 28,30,34 and 36 can follow the control signal from command block 18.Each DQ block 28,30,34 and 36 Including data input/output buffer, these buffers are controlled by the control signal of DQ control block 32.28,30,34 and of DQ block 36 are assigned to different byte numbers.
For example, memory supports the interface of 32 I/O read-write data, i.e. the data of 4 bytes, each byte has accordingly DQ block is supported.DQ block 28 provides circuit relevant to the position 24-31 in 4 character data bytes and pad.DQ block 30 provides and 4 words Save the relevant circuit in the position 8-15 and pad in data word.DQ block 34 provides circuit relevant to the position 0-7 in 4 byte datas And pad.DQ block 36 provides circuit relevant to the position the 16-23 of 4 character data bytes and pad.Particularly, DQ block 28,30,34 It include the I/O interlock circuits of each byte with 36, including output driver, pad, power supply and output control logic, other With the respective syllable dependent circuit of DQ block.
Therefore, each DQ block 28,30,34 and 36 is responsible for a byte of 4 byte words.The physics of DQ block 28,30,34 and 36 Position is extremely important, so that power consumption minimizes.First sub-block of 0-7 memory block is connected with DQ block 28 and 30, so that any number It is only stored in the first sub-block of 0-7 memory block according to the position 24-31 and 8-15 of word.Further, since the first son of memory block 0-7 The position of block is in the top half for storing core 12, therefore DQ block 28 and 30 can also be located at the top half of DQ block 14, thus More signal wires are allowed to be directly connected to reading and writing data channel 22.Likewise, the second sub-block of 0-7 memory block is connected to DQ block 34 With 36, so that 16-23 and 0-7 of any data word are only stored in the second sub-block of 0-7 memory block.Further, since storage All in the lower half portion of storage core 12, DQ block 34 and 36 also is located at the lower half portion of DQ block 14 for the position of the second sub-block of block 0-7, To allow more signal wires to be directly connected to reading and writing data channel 24.
DQ block 14 is made of DQ block 28,30,34,36, positioned at the side of memory core 12.By the way that DQ block 14 is put On one side, symmetric design can be applied to sub-block, so that the read-write power consumption across different memory blocks is symmetrical.Usually life Enable block 10 that can be located at the opposite of DQ block 14.But it can also be seen that different configurations from the present invention.For example, DQ block 14 can be with Positioned at the upright position of current DQ block 14.Or DQ block 14 can be divided into two opposite segments, so that DQ block 28 and 30 In the side of storage core 12, and DQ block 34 and 36 is located at the other side of storage core 12.
Fig. 2 is the block diagram of another embodiment of the memory that the present invention has more memory blocks.According to the present invention, order and Address (CA) block 10 and DQ block 14 can be located relative to the different location of storage core 12.For example, can be by DQ block 14 and life The function of block 10 is enabled to merge into order and DQ block 14c, they are located at the side of storage core 12.Sub-block can be divided into two groups, and one Group is located at storage core 12a, and another group is located at storage core 12b.Since order and DQ block 14c are disposed in reading and writing data channel 22 and 24 one end, by arrangement sub-block on the column direction of storage core 12a and 12b position (or the position on line direction, Depending on reading and writing data channel 22 and 24 and the direction of DQ block 14c), memory block to order and DQ block 14c substantially has identical Distance.
Fig. 3 is of the invention with another embodiment block diagram of the memory of more storage sub-blocks.In other realities of the invention It applies in example, command block 10 and DQ block 14 can be located at the centre for storing core 12, and reading and writing data channel 22 and 24 is in command block 10 With the two sides of DQ block 14.Because DQ block 14 is placed on the center of storage core, pass through the position of arrangement sub-block in a column direction (or position on line direction, depending on reading and writing data channel 22 and 24 and the direction of the opposite storage core of DQ block 14), Memory block substantially has identical distance to DQ block 14.Sub-block can be divided into two groups, and one group is located at storage core 12a, another group of position In storage core 12b.
Fig. 4 is the additional embodiment block diagram of the memory with more memory blocks of the invention.According to the present invention, command block 10 There can be different location with DQ block 14 with facing memory core 12.For example, relative to storage core 12, DQ block 14 and command block 10 The same side can be located at.DQ block 14 can be divided into two different block DQ block 14a and DQ block 14b, and wherein DQ block 14a reception comes from The data in reading and writing data channel 22 output data to reading and writing data channel 22, and DQ block 14b receives logical from reading and writing data The data in road output data to reading and writing data channel 24.Command block 10 can with DQ block and be placed on the same side, in DQ block Between 14a and 14b.Since DQ block 14a and 14b are disposed in the one end in reading and writing data channel 22 and 24, son can be arranged Position (or the position on line direction, depending on reading and writing data channel and the side of DQ block 14a and 14b of block in a column direction To), set roughly the same for the distance of memory block to DQ block 14a and 14b.Sub-block can be divided into two groups, and one group is located at storage Core 12a, another group is located at storage core 12b.
What the people for possessing common skill in this direction as one was understood, the physics based on memory of the invention Layout Embodiment has very much.Above-described embodiment is not configured to limit in any way, is only used for prominent in present invention displaying Those of embodiment.
Fig. 5 is another embodiment of the memory with storage sub-block of the invention.In LPDDR4 storage core, deposit Storing up core, there are four different region 41a-41d.Storage core 41a-41d can be divided into two groups, one group of storage core 41a and 41b, another group of storage core 41c and 41d.For every group of storage core, there is DQ block and an order and address block, for Every group of storage core write-in data read data from every group of storage core.Especially DQ block 44a passes through reading and writing data channel 43a Reading writes data to memory core 41a;DQ block 44b writes data to storage core 41b by reading and writing data channel 43b reading;Order With address block 46a for controling and operating DQ block 44a and 44b and storage core 41a and 41b.In addition, DQ block 44c passes through number Storage core 41c is write data to according to read/write channel 43c reading;DQ block 44d writes data to memory by reading and writing data channel 43d reading Core 41d;Order and address block 46b control and operate DQ block 44c and 44d, store core 41c and 41d.As shown, this hair It is bright to can extend to any number of storage core, DQ block and order and address block.The present embodiment is only based on the present invention Memory various configurations example.
Fig. 6 is the data path of memory of the invention.Command block 10 includes that order and address block 160, row address select Device 162, block control logic 164, column address counter and latch blocks 166.Order and address about data can be input to life It enables and address block 160.The address of data is transferred to storage core 12, passes through row address selector 162, block control logic 164 Data are read or are written from storage core 12 with column address counter/latch 166.Row address selector 162, block control are patrolled Volumes 164 and column address counter/latch 166 can be by activating one of memory block 0-7 read come storing data or therefrom Access evidence, to be operated to storage core 12.
Although the present invention is described some embodiments, it is to be appreciated that the invention is not limited to These embodiments.On the contrary, the present invention is understood that and explains in its broadest sense, as claim is reflected.Cause This, these claims are construed as not only including equipment described here, method and system, every other and into one The change and modification of step will be apparent from for having the people of common skill in this respect.

Claims (7)

1. a kind of framework of dynamic RAM, including a command block, the memory of multiple memory blocks, it is characterized in that each Memory block at least has the first sub-block and the second sub-block;
It further include the channel at least two reading and writing datas (DRW), wherein the first sub-block memory block is connected at least two reading and writing datas First data read/write channel in channel, and the second sub-block memory block therein are connected at least two reading and writing data channels Second data read/write channel;
And one output and input (DQ) block, wherein at least two reading and writing data channel passes between memory core and DQ block Transmission of data, wherein command block drive command signal operation memory core, reading and writing data channel and DQ block.
2. low-power consumption dynamic RAM according to claim 1, it is characterized in that in the memory block of the memory, First the second sub-block of sub-block is located in NxN array, and wherein N is an even number, wherein the first sub-block is located at the first half of array Point, wherein the second sub-block is located at the latter half of array.
3. low-power consumption dynamic RAM according to claim 1, it is characterized in that the command block is along the 1st of array Side arrangement, wherein DQ block is arranged along the 2nd side of array, wherein the 1st side and the 2nd side are located at the opposite of array.
4. low-power consumption dynamic RAM according to claim 1, it is characterized in that the first reading and writing data channel It is routed between the first half of array, the second reading and writing data channel is routed between the latter half of array.
5. low-power consumption dynamic RAM according to claims 1 and 2, it is characterized in that the first son of the memory block Block and the second sub-block are located in array, positioned at the symmetrical position of array center position.
6. low-power consumption dynamic RAM according to claim 5, it is characterized in that the wherein data word has first Range byte, the second range byte, third range byte, the 4th range byte, wherein the first DQ block handles the first range byte, 2nd DQ block handles the second range byte, and the 3rd DQ block handles third range byte, and the 4th DQ block handles the 4th range byte, In the first reading and writing data channel be connected to the first DQ block and the 2nd DQ block, the second reading and writing data channel is connected to the 3rd DQ block and Four DQ blocks.
7. low-power consumption dynamic RAM according to claim 6 leads to it is characterized in that the data are inputted from DQ block Reading and writing data channel is crossed to some memory block, data can be divided into first group and second group of data, wherein first group of data is deposited The first sub-block in some particular memory block is stored up, second group of data is stored in the second sub-block of some particular memory block.
CN201910275295.6A 2019-04-08 2019-04-08 A kind of framework of dynamic RAM Pending CN110164487A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910275295.6A CN110164487A (en) 2019-04-08 2019-04-08 A kind of framework of dynamic RAM

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910275295.6A CN110164487A (en) 2019-04-08 2019-04-08 A kind of framework of dynamic RAM

Publications (1)

Publication Number Publication Date
CN110164487A true CN110164487A (en) 2019-08-23

Family

ID=67639146

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910275295.6A Pending CN110164487A (en) 2019-04-08 2019-04-08 A kind of framework of dynamic RAM

Country Status (1)

Country Link
CN (1) CN110164487A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111445349A (en) * 2020-03-13 2020-07-24 贵州电网有限责任公司 Hybrid data storage processing method and system suitable for energy Internet
WO2023035700A1 (en) * 2021-09-13 2023-03-16 长鑫存储技术有限公司 Memory circuit and memory

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6545936B1 (en) * 1999-09-02 2003-04-08 Korea Advanced Institute Of Science & Technology Pipeline structure of memory for high-fast row-cycle
US20040081009A1 (en) * 2002-10-29 2004-04-29 Mitsubishi Denki Kabushiki Kaisha Semiconductor memory device with power consumption reduced in non-data-access
CN105190758A (en) * 2013-06-10 2015-12-23 英特尔公司 Method for improving bandwidth in stacked memory devices
CN105706171A (en) * 2013-12-18 2016-06-22 英特尔公司 Techniques for accessing a dynamic random access memory array
CN107025922A (en) * 2016-01-06 2017-08-08 瑞萨电子株式会社 Semiconductor memory
US20170255552A1 (en) * 2016-03-03 2017-09-07 Nvidia Corporation Systems and methods for dynamic random access memory (dram) sub-channels
CN107408404A (en) * 2015-02-06 2017-11-28 美光科技公司 Device and method for storage arrangement are using the storage as programmed instruction
CN109086228A (en) * 2018-06-26 2018-12-25 深圳市安信智控科技有限公司 High-speed memory chip with multiple independent access channels

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6545936B1 (en) * 1999-09-02 2003-04-08 Korea Advanced Institute Of Science & Technology Pipeline structure of memory for high-fast row-cycle
US20040081009A1 (en) * 2002-10-29 2004-04-29 Mitsubishi Denki Kabushiki Kaisha Semiconductor memory device with power consumption reduced in non-data-access
CN105190758A (en) * 2013-06-10 2015-12-23 英特尔公司 Method for improving bandwidth in stacked memory devices
CN105706171A (en) * 2013-12-18 2016-06-22 英特尔公司 Techniques for accessing a dynamic random access memory array
CN107408404A (en) * 2015-02-06 2017-11-28 美光科技公司 Device and method for storage arrangement are using the storage as programmed instruction
CN107025922A (en) * 2016-01-06 2017-08-08 瑞萨电子株式会社 Semiconductor memory
US20170255552A1 (en) * 2016-03-03 2017-09-07 Nvidia Corporation Systems and methods for dynamic random access memory (dram) sub-channels
CN109086228A (en) * 2018-06-26 2018-12-25 深圳市安信智控科技有限公司 High-speed memory chip with multiple independent access channels

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111445349A (en) * 2020-03-13 2020-07-24 贵州电网有限责任公司 Hybrid data storage processing method and system suitable for energy Internet
CN111445349B (en) * 2020-03-13 2023-09-05 贵州电网有限责任公司 Hybrid data storage processing method and system suitable for energy Internet
WO2023035700A1 (en) * 2021-09-13 2023-03-16 长鑫存储技术有限公司 Memory circuit and memory

Similar Documents

Publication Publication Date Title
KR920013444A (en) High-capacity memory and high-speed memory semiconductor memory device
US7990798B2 (en) Integrated circuit including a memory module having a plurality of memory banks
KR960038618A (en) Memory device circuit and method of simultaneous addressing of multi-bank columns of a multi-bank memory array
JP2012155837A (en) High speed fanned out system architecture and input/output circuits for non-volatile memory
KR100890381B1 (en) Semiconductor memory device
CN102347075A (en) Semiconductor device
US9136016B2 (en) Semiconductor memory apparatus
CN110164487A (en) A kind of framework of dynamic RAM
KR20150075840A (en) Memory and memory system including the same
KR100279137B1 (en) Semiconductor memory, data reading method and writing method of semiconductor memory
KR101295655B1 (en) Test apparatus and test method
US11145381B1 (en) Memory with test function and test method thereof
KR100359360B1 (en) Semiconductor memory device
CN101071631B (en) Method and device for multiple banks read and data compression for back end test
KR100618858B1 (en) Semiconductor memory device capable of changing number of banks for refreshing when conducting refresh banks and Method of refresh thereof
KR100652415B1 (en) Semiconductor memory device capable of selecting one row activation configuration or two row activation configuration and 2 row activation method thereof
JP3872922B2 (en) Semiconductor memory device and memory embedded logic LSI
US7277977B2 (en) DRAM for high-speed data access
KR100884587B1 (en) Multi port memory device
CN107527649A (en) The storage component part and its operating method of delay with raising
KR100371476B1 (en) Semiconductor integrated circuit device
US7941617B2 (en) System and device having alternative bit organization
KR19990040441A (en) I / O circuit of high speed semiconductor memory device with reduced test time
KR100697832B1 (en) Multi-port memory and method for testing the same
KR100314129B1 (en) Semiconductor implementing bank and data input/output line architecture to reduce data input/output line loading

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20190823

RJ01 Rejection of invention patent application after publication