CN106782646B - Method for writing data back to storage device and data access system - Google Patents

Abstract

A method for writing data from a memory back to a storage device, wherein the memory includes a plurality of data units for storing data, the method comprising: for each data unit, calculating the write-back energy corresponding to writing the data back to the storage device; classifying the data units into a plurality of groups according to the write-back energy corresponding to each data unit, wherein each of the groups corresponds to a write-back energy level; selecting a target group from the groups, wherein the groups corresponding to the lower write-back energy level in the groups are given a higher probability of being selected as the target group; selecting a target data unit from the target group, and writing the data of the target data unit back to the storage device.

Description

Method and data access system for writing data back to storage device

technical field

The present invention relates to a method and a data access system for writing data back from a memory to a storage device.

Background technique

Phase-Change Memory (PCM) has been widely used in various non-volatile storage devices due to its high-speed access and data retention after power off. In practice, when writing/programming the phase-change memory, the crystallization state of the phase-change material in the memory cell can be changed by heating the electrodes so that it exhibits a resistance value corresponding to the written data. For example, a multi-level memory cell (Multi-Level Cell, MLC) of a phase-change memory can be programmed to four possible resistance value distributions to represent "11", "10", "00", and "01" respectively Wait for two digits.

However, although the multi-level memory cell of the phase-change memory can store more bits of data, it requires more energy than programming a single-level memory cell (Single-Level Cell, SLC) when programming, so that System power consumption increases.

Therefore, how to reduce the energy required for writing data into the storage device so as to reduce the overall system energy consumption is one of the topics that the industry is currently working on.

Contents of the invention

The present invention relates to a method and a data access system for writing data from internal memory back to a storage device, which allows data with lower write-back energy to have a higher probability of being written back to the storage device, thereby reducing writing to the storage device The energy consumed by the data.

According to one aspect of the present invention, a method for writing data from the internal memory back to the storage device is proposed. The internal memory includes a plurality of data units for storing data. Corresponding write-back energy; according to the write-back energy corresponding to each data unit, classify these data units into a plurality of groups, and each of these groups corresponds to a write-back energy level; select a target group from these groups, wherein Among the groups, the group corresponding to the lower write-back energy level is given a higher probability to be selected as the target group; the target data unit is selected from the target group, and the data of the target data unit is written back to the storage device.

According to an aspect of the present invention, a data access system is provided, which includes a processing unit, a memory, and a storage device. The memory includes multiple data units to store data. The storage device is coupled to the memory. The processing unit is coupled to the memory, and is used for calculating the write-back energy corresponding to writing data back to the storage device for each data unit, and classifying these data units into multiple groups according to the write-back energy corresponding to each data unit. Each group corresponds to a write-back energy level, wherein the processing unit executes a write-back procedure when the memory space is full, and the write-back procedure includes: selecting a target group from these groups, wherein the corresponding lower A group with a writeback energy level is given a higher probability to be selected as the target group; and a target data unit is selected from the target group, and data of the target data unit is written back to the storage device.

In order to have a better understanding of the above and other aspects of the present invention, preferred embodiments will be described below in detail together with the accompanying drawings.

Description of drawings

Figure 1 shows a simplified block diagram of a data access system according to an embodiment of the present invention;

FIG. 2 shows a flow chart of a method for writing data from internal memory back to a storage device according to an embodiment of the present invention;

FIG. 3 shows a schematic diagram of a data structure of a memory according to an embodiment of the present invention;

FIG. 4 shows a schematic diagram of groups corresponding to different write-back energy levels according to an embodiment of the present invention;

Fig. 5 shows the operation flowchart of the data access system according to an embodiment of the present invention;

FIG. 6 shows an operation flowchart of the writeback data selection program according to an embodiment of the present invention.

【Symbol Description】

100: Data Access System

102: Processing unit

104: memory

106: storage device

A1, A2: path

Steps: 202, 204, 206, 208, 502, 504, 506, 508, 510, 512, 514, 516, 518, 602, 604, 606, 608, 610, 612, 614, 616

302: Data counter

DU, DU11~DU1i, DU21~DU2j, DUn1~DUnk: data unit

CL: cache line

DB: dirty bits

G1～Gn: Group

LV1～LVn: write back energy level

E: maximum writeback energy

Detailed ways

Herein, some embodiments of the present invention are described in detail with reference to the accompanying drawings, but not all embodiments are shown in the drawings. Indeed, these inventions can use many different variations and are not limited to the examples herein. Rather, the present invention provides these embodiments to meet the statutory requirements of the application. The same reference numerals are used in the drawings to designate the same or similar components.

FIG. 1 shows a simplified block diagram of a data access system 100 according to an embodiment of the present invention. The data access system 100 includes a processing unit 102 , a memory 104 and a storage device 106 . The processing unit 102 is, for example, a central processing unit (Central Processing Unit, CPU), a processor, a controller or other operation control circuits. The memory 104 is coupled to the processing unit 102 . The memory 104 is, for example, the main memory of the system, which can be used by the processing unit 102 to store/temporarily store data required for operations. The memory 104 can be implemented by Dynamic Random Access Memory (Dynamic Random Access Memory, DRAM) or other forms of volatile (Volatile) memory, and can also be implemented by non-volatile (Non-Volatile) memory. The storage device 106 is coupled to the memory 104 . The storage device 106 can be realized by non-volatile memory, such as phase-change memory (Phase-Change Memory, PCM), NAND flash memory, magnetoresistive random access memory (MagnetoresistiveRandom Access Memory, MRAM), variable resistance memory ( Resistive Random Access Memory, ReRAM), etc., can store data permanently.

In terms of operation, the data access system 100 can be regarded as an asymmetric access architecture. For example, the program is first stored in the storage device 106 , and then the program is loaded into the memory 104 for execution by the processing unit 102 (as shown by path A1 ). The relevant calculation results generated during the execution of the processing unit 102 can be temporarily stored in the memory 104 . When the space of the memory 104 is full, or when certain data must be stored back into the storage device 106, the processing unit 102 will perform a write back (write back) action to write the data from the memory 104 back to the storage device 106 (as indicated by path A2 Show). Since writing data back into the storage device 106 requires a certain amount of write-back energy (for example, the energy spent on programming the phase-change memory by applying thermal energy), if the write-back energy required to perform the write-back operation can be reduced, it will be possible Effectively reduce system energy consumption.

According to an embodiment of the present invention, the processing unit 102 can first evaluate the write-back energy of each data unit in the memory 104, and then classify these data units into groups corresponding to different write-back energy levels according to their corresponding write-back energy . Afterwards, combined with the method of probability selection, the group corresponding to the low energy level has a higher probability to be selected, and then finds suitable write-back data from the selected group. In this way, most of the data selected to be written back to the storage device 106 corresponds to low write-back energy, so that the system power consumption can be effectively reduced.

FIG. 2 shows a flowchart of a method for writing data from the memory 104 back to the storage device 106 according to an embodiment of the invention. In step 202 , the processing unit 102 calculates write-back energy corresponding to writing data back to the storage device 106 for each data unit of the memory 104 . The data unit may be a memory page or other data management units, depending on the access architecture of the system.

In one embodiment, the write-back energy corresponding to each data unit can be determined by counting the number of high-energy level data in each data unit. The high-level data refers to data that requires more energy to be programmed to reach a target data level, for example, corresponding to one or more specific bit patterns. For example, when programming a storage cell of a phase-change memory, the first bit of data can be written in a manner similar to programming a single-level cell (Single-Level Cell, SLC) to form the corresponding data " 1" (or "11"), "0" (or "00") resistance value distribution. If the second bit is to be written later, more precise programming control is required to further separate the resistance value distribution corresponding to the data "10" and "01". Therefore, the energy required to write data "10", "01" will be significantly higher than the energy required to write data "11", "00". Therefore, in this example, the data with the bit pattern of "10" and "01" will be regarded as high-level data, and the processing unit 102 can count how many data with the bit pattern of "10" and "01" contained in the data unit to estimate the writeback energy corresponding to the data unit.

It can be understood that the present invention is not limited to the above-mentioned embodiments, and the data corresponding to different memory cell distributions can be adjusted according to different designs, so that the bit patterns corresponding to high-level data are different. In other examples, the processing unit 102 may further consider other factors to calculate the write-back energy corresponding to the data unit. Relevant descriptions will be described in conjunction with FIG. 3 .

In step 204, the processing unit 102 may classify the data units into a plurality of groups according to the write-back energy corresponding to each data unit, wherein each of the groups corresponds to a write-back energy level. For example, the processing unit 102 can classify the data units whose write-back energy is in the first write-back energy range (such as 50 Joules to 150 Joules) into the first group, and put the write-back energy in the second write-back energy range The data units (for example, 150 Joules to 250 Joules) are classified into the second group, and so on until all the data units are classified into appropriate groups. At this time, the first group corresponds to a lower write-back energy level, which includes data units whose write-back energy falls between 50 Joules and 150 Joules, and the second group corresponds to a higher write-back energy level. Writeback energy level, which includes data units whose writeback energy falls between 150 Joules and 250 Joules. In other words, the so-called group corresponding to a relatively low write-back energy level means that the write-back energy of the data units contained in the group falls within a relatively low write-back energy range; and the so-called relatively high write-back energy level A group means that the writeback energy of the data units included in the group falls within a relatively high writeback energy range.

In step 206 , the processing unit 102 selects a target group from the groups, wherein a group corresponding to a lower writeback energy level among the groups is assigned a higher probability to be selected as the target group. For example, if the processing unit 102 selects the target group in a random manner, the number of samples corresponding to groups with low writeback energy levels may be increased in the randomly selected sample space, and the number of samples corresponding to groups with high writeback energy levels may be decreased. number, thereby increasing the probability that a group with a low writeback energy level will be selected as the target group. For example, suppose there are three groups A, B, and C, and the corresponding write-back energy ranges are, for example, 0-100 Joules, 200-300 Joules, and 300-400 Joules respectively, and the three groups A, B , C are given different weights, for example, 4, 2, 1 respectively. At this time, the probability of selecting group A is 4/7, the probability of selecting group B is 2/7, and the probability of selecting group The probability of C is 1/7. Therefore, group A with relatively low writeback energy will have a higher probability of being picked than other groups B, C. In this way, the storage device 106 will have a higher probability of being written back with low write-back energy data, thereby reducing the energy consumed for the write-back operation on the storage device 106 .

In one embodiment, the weight setting of the probabilistic selection group can start from the group corresponding to the highest writeback energy level, specifying its weight as 1, the second highest as 2, and the third highest as 4, so that By analogy, these weights are then added together as the denominator, and the numerator is the weight of the corresponding group (that is, the weight of the corresponding group divided by the sum of the weights), so as to obtain the probability of being selected corresponding to the group. In this way, groups with relatively low writeback energy will have a higher probability of being selected. It can be understood that the setting of the weight values adopted in the above-mentioned embodiments is only for illustration purposes, and is not intended to limit the present invention. In other embodiments, the setting of the weight value can also be adjusted according to different applications.

In addition, through the distribution of probability, the influence of cold data (Cold Data) and hot data (Hot Data) on energy consumption can be eliminated. Furthermore, although some data has low write-back energy, its write-back frequency is high (hot data), so that the overall write-back energy is still high. Through the probability selection method, some data (cold data) with high write-back energy but low write-back frequency will have the opportunity to be selected for write-back, thereby slowing down the impact of cold and hot data on energy consumption.

In step 208 , the processing unit 102 selects the target data unit from the selected target group, and writes the data of the target data unit back to the storage device 106 . The processing unit 102 may select the target data unit based on a random method or a certain rule. Since each data unit has been classified, and the group corresponding to the low writeback energy level is given a higher probability of being selected, the processing unit 102 does not need to compare the writeback energy of each data unit one by one to find out the relatively low writeback energy level. Write back the energy data. Therefore, the processing unit 102 can quickly select the data to be written back to the storage device 106 in this way, thereby shortening the overall programming time.

In one embodiment, after the processing unit 102 performs the write operation on the memory 104, it will manage and classify the relevant data units as in steps 202 and 204, and when the space of the memory 104 is full, perform the operations as in step 206 And the write-back procedure of 208.

FIG. 3 shows a schematic diagram of a data structure of a memory according to an embodiment of the present invention. For convenience of description, the memory described here takes the memory 104 in FIG. 1 as an example, but this does not limit the present invention.

In the example of FIG. 3 , the memory 104 includes a plurality of data units DU, and each data unit DU includes a plurality of cache lines CL. The data unit DU is, for example, a data page, and its size is, for example, 4K bytes (Byte) or 4M bytes, and the size of a cache line CL is, for example, 64 bytes or 128 bytes, depending on the actual application.

Each data unit DU is assigned a data counter 302 respectively, and each data counter 302 is used to calculate how many high-energy-consuming data (such as data whose bit patterns are "10" and "01") in the corresponding data unit DU, for The processing unit 102 estimates the write-back energy corresponding to each data unit DU.

In one embodiment, the data counter 302 does not need to count the amount of high-power consumption data contained in all the cache lines CL in the corresponding data unit DU, but only needs to count the high-power consumption data in the cache lines CL that have modified data. The amount of data can be used. Furthermore, each cache line CL is assigned a dirty bit DB to indicate whether data has been modified/written to it. When a cache line CL is marked by the dirty bit DB (for example, the value of the dirty bit DB=1), it means that data has been modified on the cache line CL. On the contrary, when the cache line CL is not marked by the dirty bit DB (for example, the value of the dirty bit DB=0), it means that the cache line CL has not been modified data. Generally speaking, only modified data may need to be written back to the storage device 106, so the data counter 302 can count the high-level data contained in the cache line CL marked by the dirty bit DB in each data unit DU. The number is used for the processing unit 102 to calculate the write-back energy corresponding to each data unit DU. For example, the processing unit 102 can calculate the writeback energy corresponding to the data unit DU based on the following formula:

(Formula 1)

Wherein Epf represents the write-back energy corresponding to a data unit DU, Cpc represents the counting result of the data counter 302, Cdirty represents the number of cache lines CL marked by the dirty bit DB in the data unit DU, and Ncl represents the number of cache lines CL in a data unit DU. Ehe represents the average write-back energy of writing high-level data (such as data "01" or "10"), and Ele represents the average write-back energy of writing low-level data (such as data "00" or "11") energy.

Furthermore, the item Ncl×Cdirty is the number of bits contained in all the cache lines CL marked by the dirty bit DB, and for a multi-level memory cell (MLC), since each write operation is to write two Therefore, in (Equation 1), divide the term Ncl×Cdirty by 2 to calculate the actual writing action. In addition, the reason why the term Cpc×Ehe is not divided by 2 in (Formula 1) is that Cpc itself records the number of high-level data (such as data "01" or "10"), rather than recording how many The ones digit, so there is no need to divide by 2.

It can be understood that the present invention is not limited to the above-mentioned embodiments, and any evaluation of the writeback energy of a corresponding data unit based on the amount of high-energy level data in the data unit falls within the spirit of the present invention.

FIG. 4 shows a schematic diagram of groups corresponding to different writeback energy levels according to an embodiment of the invention. In the example shown in FIG. 4 , there are n groups G1 -Gn, where n is a positive integer. The groups G1-Gn respectively correspond to the write-back energy levels LV1-LVn, and are arranged sequentially from low write-back energy to high write-back energy. Data groups corresponding to the same writeback energy level/writeback energy range are classified into the same group. As shown in Figure 4, data units DU11~DU1i corresponding to writeback energy level LV1 are classified into group G1; data units DU21~DU2j corresponding to writeback energy level LV2 are classified into group G2; The data units DUnl˜DUnk of the level LVn are classified into the group Gn, where i, j, k are positive integers. For example, assuming that the write-back energy level LV1 indicates that the write-back energy is 100 Joules to 200 Joules, and the write-back energy level LV2 indicates that the write-back energy is 200 Joules to 400 Joules, at this time, if the first data unit , the second data unit, and the third data unit are written back to the storage device 106 to consume 130 joules, 180 joules, and 300 joules of energy respectively, then the first and second data units will be classified into the same group G1, and the second Three data units will be classified into group G2.

In an embodiment, the write-back energy intervals corresponding to the groups G1 -Gn may be equally divided. For example, if the maximum write-back energy is E, the interval widths of the write-back energy intervals corresponding to the groups G1 -Gn may all be set as E/n. At this time, group G1 corresponds to 0 to The writeback energy range of the group G2 corresponds to to The writeback energy interval of the group G3 corresponds to to The writeback energy range of , and so on.

In yet another embodiment, the write-back energy intervals corresponding to the groups G1 -Gn may be divided unevenly. For example, if the maximum write-back energy is E, the write-back energy intervals corresponding to the groups G1 -Gn can be divided based on the power of a specific base Q. Taking base Q equal to 2 as an example, group G1 corresponds to 0 to The writeback energy range of the group G2 corresponds to to The writeback energy range of the group G3 corresponds to to The writeback energy range of , and so on. If the groups G1-Gn are arranged in order from low to high according to the write-back energy level, the write-back energy interval width corresponding to the low write-back energy group is Q of the write-back energy interval width corresponding to the high write-back energy group ^R times, where Q represents the base of the maximum write-back energy E divided by the power, R represents the ordinal difference between the two groups, and both Q and R are positive integers. Taking the aforementioned example as an example, when the base Q is equal to 2, the writeback energy interval width of group G1 (corresponding to ordinal number 1) is 2 ^{(3 -1)} times.

It can be understood that the present invention is not limited to the above examples. The writeback energy range corresponding to each group may also be divided based on other rules, depending on different applications.

FIG. 5 shows an operation flowchart of the data access system according to an embodiment of the present invention. For convenience of description, the data access system described here takes the data access system 100 in FIG. 1 as an example, but this does not limit the present invention.

In step 502 , the processing unit 102 first determines whether the data unit (such as a data page) to be accessed exists in the memory 104 . If yes, continue to step 504, and the processing unit 102 will determine whether to perform a read or write operation on the data unit. If not, then proceed to step 516, and the processing unit 102 then determines whether the memory 104 is full. If yes, enter step 506 , the processing unit 102 executes the write-back data selection program to select the appropriate write-back data from the memory 104 and write it back into the storage device 106 . An exemplary detailed flow of the selection program will be described with reference to FIG. 6 .

If the result of step 516 is negative, that is, the space of the memory 104 has not been filled, the processing unit 102 will execute step 518 to create the data unit to be accessed in the memory 104 . For example, the processing unit 102 can read from the storage device 106 or directly create an empty memory page on the memory 104 for reading and writing.

In step 510, when the processing unit 102 determines that a read operation is to be performed on the data unit, the processing unit 102 reads data from the data unit.

In step 508, when the processing unit 102 determines that a write operation is to be performed on the data unit, the processing unit 102 will write data to the data unit. Then, as shown in steps 512 and 514, for the written data units, the processing unit 102 will calculate their corresponding write-back energy, and classify them into corresponding groups according to the obtained write-back energy, so that they can be written later. Back to the selection of data.

FIG. 6 shows an operation flowchart of the writeback data selection program according to an embodiment of the present invention. The operation flow chart can be, for example, a detailed flow chart of step 506 in FIG. 5 , but the invention is not limited thereto.

In step 602 , the processing unit 102 first initializes a value of a loop counter, for example, resets it to 0.

In step 604, the processing unit 102 randomly selects groups. For example, each group is given a different probability of being selected, so that a group with a low energy level is easier to be selected than a group with a high energy level.

In step 606 , the processing unit 102 adjusts the value of the loop counter to count the number of times the processing unit 102 randomly selects groups. For example, the loop counter value can be adjusted based on the following formula:

i=i+1 (Formula 2)

where i represents the value of the loop counter.

In step 608, the processing unit 102 determines whether there are any data units in the selected group. If not, it means that the processing unit 102 needs to reselect the group. As shown in step 612 , the processing unit 102 will determine whether the value of the loop counter exceeds a predetermined threshold, and when it is determined that the value of the loop counter has not exceeded the predetermined threshold, execute step 604 again to reselect groups. Conversely, when the value of the loop counter exceeds the predetermined threshold, it means that the processing unit 102 has performed multiple random selection procedures and still cannot select a valid group (that is, a group with data units stored therein). At this time, the processing unit 102 Step 614 will be executed to search for valid groups in a fixed order, for example, starting from the group corresponding to the lowest writeback energy level to find the first group with data units.

The selected effective group can be regarded as the target group. After the target group is selected, the processing unit 102 selects the target data unit from the selected target group and writes the target data unit back to the storage device 106 , as shown in steps 610 and 616 . After completing the write-back data selection procedure shown in FIG. 6 , the processing unit 102 may, for example, proceed to step 518 in FIG. 5 to perform subsequent related operations on the memory 104 .

To sum up, the method and data access system for writing data back from the memory to the storage device proposed by the present invention can first estimate the write-back energy of each data unit in the memory, and then write these data units according to their corresponding Writeback energies are grouped into groups corresponding to different writeback energy levels. Afterwards, combined with the method of probability selection, the group corresponding to the low energy level has a higher probability to be selected, and then finds suitable write-back data from the selected group. In this way, not only can the system energy consumption be effectively reduced, but also the speed of selecting suitable write-back data can be increased, thereby shortening the overall programming time.

Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be defined by the appended claims.

Claims (10)

1. a kind of method that data are write back storage device from memory, which is characterized in that the memory include multiple data cells with Data are stored, this method comprises:

For the respectively data cell, calculating, which writes back data, writes back energy corresponding to the storage device；

Energy is write back according to corresponding to the respectively data cell, these data cells are categorized into multiple groups, these groups are each Energy grade is write back from corresponding one；

A target group is selected from these groups, wherein corresponded in these groups the lower group for writing back energy grade than it is corresponding compared with The group that height writes back energy grade is endowed higher probability and is chosen as the target group；And

A target data unit is selected from the target group, and the data of the target data unit are write back into the storage device.

2. the method as described in claim 1, which is characterized in that further include:

The number of high energy order evidence in the respectively data cell is counted, to determine to write back energy corresponding to the respectively data cell；

Wherein the high energy order is according to one or more corresponding certain bits patterns.

3. the method as described in claim 1, which is characterized in that wherein this these data cell respectively includes a plurality of cache line, This method further include:

The number for the high energy order evidence for being included by dirty position (dirty bit) cache line marked in the respectively data cell is counted, To determine to write back energy corresponding to the respectively data cell；

Wherein the high energy order is according to being one or more corresponding certain bits patterns.

4. the method as described in claim 1, which is characterized in that further include:

Energy will be write back in this these data cell be located at one first data cell for writing back energy section sort out to this these group One first group in group；And

Energy will be write back in this these data cell be located at one second data cell for writing back energy section sort out to these groups In one second group；

Wherein first interval width for writing back energy section is equal to second interval width for writing back energy section.

5. the method as described in claim 1, which is characterized in that further include:

Energy will be write back in this these data cell be located at one first data cell for writing back energy section sort out to these groups In one first group；And

Energy will be write back in these data cells be located at one second data cell for writing back energy section sort out into these groups One second group；

Wherein this first writes back energy section and this second writes back power side of the energy section based on a specific truth of a matter to one most Capitalization resilience amount is divided and is generated.

6. a kind of data access arrangement characterized by comprising

One memory, including multiple data cells are to store data；

One storage device couples the memory；And

One processing unit couples the memory, to for the respectively data cell, calculating writes back data corresponding to the storage device Write back energy, and write back energy according to corresponding to the respectively data cell, these data cells be categorized into multiple groups, this A little groups respectively correspond to one and write back energy grade, and wherein the processing unit executes one when the memory headroom has been expired and writes back program, This writes back program

A target group is selected from these groups, wherein corresponded in these groups the lower group for writing back energy grade than it is corresponding compared with The group that height writes back energy grade is endowed higher probability and is chosen as the target group；And

A target data unit is selected from the target group, and the data of the target data unit are write back into the storage device.

7. data access arrangement as claimed in claim 6, which is characterized in that further include:

Multiple data counters, to count the number of high energy order evidence in these data cells, for processing unit calculating Respectively energy is write back corresponding to the data cell；

Wherein the high energy order is according to one or more corresponding certain bits patterns.

8. data access arrangement as claimed in claim 6, which is characterized in that wherein these data cells respectively include a plurality of fast Line taking, the data access arrangement further include:

Multiple data counters, to count in the respectively data cell, cache line by dirty position (dirty bit) label includes High energy order evidence number, respectively write back energy corresponding to the data cell so that the processing unit calculates；

Wherein the high energy order is according to one or more corresponding certain bits patterns.

9. data access arrangement as claimed in claim 6, which is characterized in that wherein the processing unit will be in these data cells Write back energy and be located at one first data cell for writing back energy section and sort out one first group into these groups, and by these Energy is write back in data cell be located at one second data cell for writing back energy section sort out one second group into these groups Group；

Wherein first interval width for writing back energy section is equal to second interval width for writing back energy section.

10. data access arrangement as claimed in claim 6, which is characterized in that wherein the processing unit is by these data cells In write back energy and be located at one first data cell for writing back energy section and sort out one first group into these groups, and by this Write back in a little data cells energy be located at one second data cell for writing back energy section sort out into these groups one second Group；

Wherein this first writes back energy section and this second writes back power side of the energy section based on a specific truth of a matter to one most Capitalization resilience amount is divided and is generated.