Estimating the initial state of a resistive element

Abstract

Estimating the Initial State of a Resistive Element. This description relates to a method for estimating the resistive state in which a memory cell is programmed. The method comprises: a) a first measurement, by a measuring circuit connected to the cell, of a first representative value of the cell's resistance; b) at least one second measurement of at least one second representative value of the cell's resistance; c) a comparison, by a comparison circuit, of the first and at least one second representative values ​​of the resistance; and d) based on the comparison, an estimation of the cell's resistive state. The estimation is either a high-resistance state if at least one second value shows an increase compared to the first representative value, or a low-resistance state if at least one second value shows a decrease compared to the first value. Figure for the abstract: Fig. 6A

Description

Title of the invention: Estimation of P initial state of a resistive element Technical field

[0001] The present description relates generally to resistive non-volatile memories, and in particular to the estimation of the resistance states in which each resistive element of the memory was initially programmed. Prior art

[0002] Resistive volatile memories comprise a plurality of resistive elements. Each resistive element is initially programmed to a high resistance state (HRS) or a low resistance state (LRS).

[0003] However, once programmed, a resistive element undergoes relaxation effects, causing fluctuations in the resistance to which it was initially programmed. Thus, an element programmed to a high resistance state may see its resistance decrease and become in a low resistance state, and vice versa for an element programmed to a low resistance state.

[0004] Although the relaxation effects stabilize after a certain time, fluctuations in resistance still occur.

[0005] There is a need to estimate the state in which a resistive element of a resistive non-volatile memory was initially programmed. Summary of the invention

[0006] One embodiment provides a method for estimating the resistive state in which a cell of a non-volatile resistive memory is programmed, the method comprising: a) a first measurement, by a measuring circuit connected to the memory cell, of a first value representative of the resistance of the cell; b) at least a second measurement, by the measuring circuit, of at least a second value representative of the resistance of the same memory cell; c) comparing, by a comparison circuit, the first and at least one second value representative of the resistance; and (d) based on the comparison, estimating the resistive state of the cell, the estimate being either an estimate of a high resistance state if the at least one second representative value exhibits an increase relative to the first representative value of the resistance, or an estimate of a low resistance state if the at least one second representative value exhibits a decrease relative to the first representative value.

[0007] According to one embodiment, the above method further comprises: - before carrying out step a), a first application, at a first instant, of a current of predefined intensity or of a predefined voltage to the cell of the resistive memory, the first measurement being carried out on the basis of the first application; and - following the performance of step a) and before the performance of step b), at least one second application, at at least one second instant, subsequent to the first instant, of the current or a predefined voltage to the resistive element, the at least one second measurement being carried out on the basis of the at least one second application.

[0008] According to one embodiment, the above method further comprises, following the estimation of the state: - reprogramming the cell to the estimated state.

[0009] According to one embodiment, the above method further comprises, following the estimation of the state: - storage in a memory, in association with the cell, of a value indicating the estimated state.

[0010] According to one embodiment, the above method further comprises following the measurement of the first representative value: - the comparison of the first representative value with a reference interval; and - if the first representative value is outside the reference interval, the process is repeated in a new implementation of step a).

[0011] According to one embodiment, the above method further comprises, before the comparison: - the conversion, by a converter, of the first and at least one second representative values ​​into a first and at least one second digital values; - storing, in association with the cell, the first and at least one second digital values ​​in a memory.

[0012] According to one embodiment, the memory is a shift register.

[0013] According to one embodiment, the comparison is carried out on the basis of the values ​​of voltage stored by a first and at least one second capacitor configured to respectively store the first and at least one second representative values.

[0014] According to one embodiment, the estimated state is a high resistance state if during the comparison, it is determined that the first representative value is less than one of the at least one second representative value and in which the estimated state is a low resistance state if during the comparison, it is determined that the first representative value is greater than one of the at least one second representative value.

[0015] According to one embodiment, the estimated state is a high resistance state if during the comparison, it is determined that the first representative value is lower than the average of the at least one second representative values.

[0016] According to one embodiment, the at least one second representative value comprises a first second value and a second second value, the second second value being measured at a time subsequent to the second time, and wherein the estimated state is a high resistance state if during the comparison, it is determined that the average of the differences between the first second and the second second representative values ​​and between the first second and the first representative values ​​is positive and wherein the estimated state is a low resistance state if during the comparison, it is determined that the average of the differences between the first second and the second second representative values ​​and the second representative values ​​and between the first second and the first representative values ​​is negative.

[0017] According to one embodiment, the second instant is distant from the first instant by a duration of at least 3 seconds, and for example at least one minute.

[0018] According to one embodiment, the first and at least one second representative resistance values ​​are voltage or current values ​​representative of the resistance of the cell.

[0019] One embodiment provides a circuit comprising a non-volatile resistive memory comprising a cell, programmed to one of a plurality of states, the circuit further comprising: ; - a measuring circuit configured to measure a first value representative of the resistance of the cell and at least one second representative value; - a comparison circuit configured to compare the first representative value with the at least one second representative value to estimate the state of the cell, the estimate being an estimate of a high resistance state if the at least one second representative value exhibits an increase relative to the first representative value of the resistance or of a low resistance state if the at least one second representative value exhibits a decrease relative to the first representative value.

[0020] According to one embodiment, the above circuit further comprises a current source configured to apply a current of predefined intensity, or a voltage source configured to apply a predefined voltage, to the cell at a first instant and then at at least a second instant, subsequent to the first instant, the measurement of the first representative value being carried out on the basis of the application at the first instant and the measurement of one or more representative values ​​being carried out on the basis of the application to the at least second instant.

[0021] According to one embodiment, the non-volatile memory is a filamentary type memory, for example a resistive random access memory (RRAM).

[0022] According to one embodiment, the non-volatile memory is a phase change memory, or an oxide-based resistive memory, or a programmable metallization cell. Brief description of the drawings

[0023] These characteristics and advantages, as well as others, will be explained in detail in the following description of particular embodiments given without limitation in relation to the attached figures among which:

[0024] [Fig.l] is a graph illustrating fluctuations in the resistance of resistive elements programmed into high resistance states or low resistance states;

[0025] [Fig.2A] is a graph illustrating fluctuations in the resistance of resistive elements around a threshold interval;

[0026] [Fig.2B] illustrates the resistance variations around the threshold interval;

[0027] [Fig.3] is a graph illustrating the evolution over time of the resistance of resistive elements;

[0028] [Fig.4A] is a graph illustrating the deviation in fluctuations of programmed resistive elements in a high resistance state;

[0029] [Fig.4B] is a graph illustrating the deviation in the fluctuations of resistive elements programmed in a low resistance state;

[0030] [Fig.5] is a graph illustrating the average fluctuations of resistive elements programmed into high or low resistance states;

[0031] [Fig.6A] is a graph illustrating two measurements;

[0032] [Fig.6B] is a graph illustrating several measurements; and

[0033] [Fig.7] is a block diagram illustrating a circuit according to an embodiment of the present description. Description of the embodiments

[0034] The same elements have been designated by the same references in the different figures. In particular, the structural and / or functional elements common to the different embodiments may have the same references and may have identical structural, dimensional and material properties.

[0035] For the sake of clarity, only the steps and elements useful for understanding the described embodiments have been shown and are detailed. In particular, the tech nologies of resistive memories, as well as the reprogramming of resistive elements, are known to the person skilled in the art. In particular, the reprogramming of the cells of a filamentary memory to a state of high, and / or low, resistance is known to the person skilled in the art.

[0036] Unless otherwise specified, when referring to two elements connected to each other, this means directly connected without intermediate elements other than conductors, and when referring to two elements connected (in English "coupled") to each other, this means that these two elements can be connected or be connected by means of one or more other elements.

[0037] In the following description, when reference is made to absolute position qualifiers, such as the terms "front", "back", "top", "bottom", "left", "right", etc., or relative position qualifiers, such as the terms "above", "below", "upper", "lower", etc., or to orientation qualifiers, such as the terms "horizontal", "vertical", etc., reference is made unless otherwise specified to the orientation of the figures.

[0038] Unless otherwise specified, the expressions "about", "approximately", "substantially", and "of the order of" mean to within 10%, preferably to within 5%.

[0039] [Fig. 1] is a graph 100 illustrating the fluctuations in the resistance of resistive elements programmed in high resistance states and in low resistance states. More particularly, curves 102 and 104 were obtained following experiments. For example, the experiments carried out measure the value of the resistance of resistive elements, for example following a measurement of the output voltage of the elements. The resistive elements are for example included in cells of a non-volatile resistive memory ReRAM (Resistive Random Access Memory). For example, the non-volatile memory is a filamentary memory. In other examples, the non-volatile memory is a phase change memory (PCM), an OxRAM (Oxide Random Access Memory) type memory, a programmable metallization cell, etc.

[0040] Curve 102 illustrates the distribution of resistance measurements ( / ?[□]) of resistive elements initially programmed to a high resistance state (HRS - "High Resistance State"). More particularly, the measurements are carried out a period of time after the programming of the resistive elements. For example, the measurements were carried out 6 seconds after the initial programming.

[0041] Curve 104 illustrates the distribution of resistance measurements of resistive elements initially programmed to a low resistance state (HRS - "High Resistance State"). More particularly, the measurements are made a period of time after the programming of the resistive elements. For example, the measurements were made 6 seconds after the initial programming.

[0042] The y-axis of graph 100 represents the quantiles (Q) for the measurements made. In particular, the quantiles represented are relative to a normal distribution, theoretical for the measurements made. For example, the resistance value corresponding to the quantile equal to 0 is the median resistance value. Similarly, the set of resistance values ​​located between the quantile values ​​-1 and 1 represents 68.27% of the total number of values. In other words, the unit Q is the standard deviation of the normal distribution of the resistance values.

[0043] Generally speaking, the resistance values ​​of the programmed cells in the high and low resistance states follow log-normal laws. In addition, the resistances taken by elements programmed in a high resistance state are more dispersed than the resistances taken by the elements programmed in low resistance states. Indeed, the resistance values ​​of the LRS elements illustrated in FIG. 1 are in an interval going from 1000 ohms to approximately 20,000 ohms while the resistances of the HRS elements are spread in an interval going from 10,000 ohms to H)7 ohms.

[0044] An interval 106 represents a range of resistance values ​​that can be assumed by resistive elements initially programmed in the high resistance state and in the low resistance state. The interval 106 extends from a resistance value of RtM ohms to a value Rthz ohms. For example, Rthi = 11000 and Rtk2 ~ 25,000. More generally, Rth\ is a value in the range 8,000 ohms to 20,000 ohms, and Rth2 is a value in the range 20,000 ohms to 50,000 ohms. The preceding values ​​R^ and Rthz are given by way of example and may differ, for example, from one type of memory to another.

[0045] Furthermore, following their initial programming, the resistance values ​​of the LRS and HRS elements undergo a relaxation effect and fluctuate. Even after the resistance distributions of the HRS and LRS elements have stabilized, the resistance of each cell continues to fluctuate individually.

[0046] In the case of filamentary memories, the relaxation effect comes from the dynamic stabilization of the filament, which modifies the cellular resistance. Indeed, atoms, for example oxygen atoms, can move and recombine for a certain period of time and until stabilization is reached.

[0047] Fluctuations before and after stabilization do not change the distributions of resistance values ​​for each HRS and LRS state. However, when values ​​fluctuate around and within the 106 range, it is difficult to distinguish an HRS element from an LRS element.

[0048] [Fig.2A] is a graph illustrating fluctuations in the resistance of resistive elements around the interval 106.

[0049] [Fig.2B] illustrates the resistance variations around the interval 106.

[0050] The y-axis of the graph illustrated in [Fig.2A] represents the quantiles, as described in relation to [Fig.l], relative to a theoretical normal distribution for measurements made. The two inclined segments represent, respectively, the theoretical normal distributions for the LRS and HRS states. For example, the abscissa of the graph illustrated in [Fig.2A] is in logarithmic scale.

[0051] By way of example, a resistance value 200 represents the resistance value of an element initially programmed in an LRS state and having fluctuated until it belongs to the threshold interval 106. The resistance of this element will then generally decrease, or decrease, in order to exit the interval 106. Similarly, a resistance value 202 represents the resistance value of an element initially programmed in an HRS state and having fluctuated until it belongs to the threshold interval 106. The resistance of this element will then generally increase, or increase, in order to exit the interval 106. Generally speaking, the resistance of an HRS or LRS element, when it enters the interval 106, will tend to return to a value specific to its initial state.

[0052] Outside the interval 106, the LRS elements, for example having a resistance 204, generally see their resistance increase. The HRS elements, for example having a resistance 206, generally see their resistance decrease. These growths and decreases outside the interval 106 are trends. Indeed, the resistance of HRS elements will sometimes fluctuate and increase, as is the case for example for the resistance 208.

[0053] Thus, when a resistive element HRS or LRS having its resistance in the interval 106 will see its resistance leave the interval 106, another resistive element programmed in the same state will see its resistance fluctuate and enter the interval 106.

[0054] [Fig.3] is a graph 300 illustrating the evolution over time of the resistance of HRS resistive elements.

[0055] In particular, graph 300 illustrates the evolution of the logarithm of the resistance value (Jog(R)) as a function of time (T[s]). Graph 300 comprises curves 302, 304 and 306 dividing the evolution of the resistances into terciles. In other words, each curve 302, 304 and 306 illustrates the evolution of the resistance of the same number of resistive elements. Curve 302 illustrates the evolution of the resistances of the upper tercile, that is to say of the third of the resistive elements having the highest resistance values. Curve 304 illustrates the evolution of the resistances of the middle tercile. Curve 306 illustrates the evolution of the resistances of the lower tercile, that is to say of the third of the resistive elements having the lowest resistance values.

[0056] The resistances of the upper tercile tend to decrease over time, while the resistances of the middle and lower terciles tend to increase. The tercile means for the HRS and LRS states tend to move closer together. This ratio closely is carried out with an increasing deviation. In fact, each element whose resistance will increase in a distribution is replaced by another element whose resistance will decrease.

[0057] [Fig.4A] is a graph 400 illustrating the deviation in the fluctuations of resistive elements programmed in a high resistance state.

[0058] [Fig.4B] is a graph 402 illustrating the deviation in the fluctuations of resistive elements programmed in a low resistance state.

[0059] In particular, the graphs 400 and 402 comprise point clouds 400' and 402'. Each point of the clouds 400' and 402' has as its abscissa coordinate the logarithm of the resistance of a resistive element at an instant f^Og^R) (fj ). For example, the instant li occurs a few seconds, for example 6 seconds, after the programming of the resistive element. Each point of the clouds 400 and 402 has as its ordinate coordinate the difference of the logarithms of the resistance of the resistive element between an instant and the instant t] (log(R) (o) — log(R) (^)). For example, the instant is later than the instant U and occurs, for example, several tens of minutes after the programming of the element, for example, one hour after the programming. In particular, cloud 400' illustrates the temporal variation of the resistance of resistive elements programmed in an HRS state. Cloud 402' illustrates the temporal variation of the resistance of resistive elements programmed in an LRS state..

[0060] A frame 404 shows a negative drift in the time variation of the resistance of resistive elements programmed in an HRS state, which illustrates the tendency of the resistances of HRS elements to increase. Similarly, a frame 406 shows a positive drift in the time variation of the resistance of resistive elements programmed in an LRS state, which illustrates the tendency of the resistances of LRS elements to decrease. These trends are only significant for elements whose resistances are at values ​​in frames 404 and / or 406.

[0061] Figure 5 is a graph 500 illustrating the average fluctuations of resistive elements programmed in high or low resistance states. More particularly, the graph 500 comprises a curve 502 and a curve 504. The curves 502 and 504 are respectively obtained by successive measurements of the resistance of a plurality of resistive elements HRS and LRS when in the threshold interval 106. The curves 502 and 504 respectively illustrate the ratio (RATIO) of resistive elements whose resistance value has increased, respectively decreased. For example, for each instant, the increase, or the decrease is obtained by comparison between the resistance value measured at instant 1 and the average of the resistance values ​​measured at several instants in the 2 minutes following instant C

[0062] The constant curves 506 and 508 respectively illustrate the average ratio of HRS elements whose resistance increases and the average ratio of LRS elements whose resistance decreases.

[0063] [Fig.6A] is a graph illustrating two measurements. As an example, [Fig.6A] illustrates an example of measurements making it possible to verify whether a resistance belonging to the threshold interval 106 tends to increase or decrease.

[0064] The example illustrated in Figure 6A shows a measurement Re^Tb carried out at a time lj. For example, at time ?i, the resistance of the resistive element tested is equal to ohms. For example, the value belongs to the threshold interval 106. Another measurement Re{T2 of the resistance of the resistive element tested is for example carried out at a time U, after the time. For example, time ?2 is distant from the time by one or more seconds, for example at least 3 seconds, by one or more minutes, for example at least 1 minute, or by at least one hour. For example, the value of the resistance at time ^2 is R2 ohms. In the case where R2 is greater than Ry, as illustrated in Figure 6A, it will be agreed that the value of the resistance has increased. Otherwise, that is, if R\ > R2, we agree that the resistance value has decreased. In one example, the value Æ2 does not belong to the threshold interval 106.

[0065] Figure 6B is a graph illustrating several measurements Re{F^ Re^T2, Re{T^ R^\T^ Re}Tn, \ as an example, [Fig.6B] illustrates another example of measurements making it possible to verify whether a resistance belonging to the threshold interval 106 tends to increase or decrease.

[0066] In the example illustrated in Figure 6B, n- 1, n being an integer, for example less than 1000, other measurements are carried out at times subsequent to the time fj. For example, the consecutive measurements are carried out in a relatively short period of time, for example during a period of time between 1 ms and 1 s. The number of total measurements that can be carried out then depends on the measurement time as well as the processing time of the measurements carried out. For example, the integer n is between 10 and 1000. For example, the other measurements are carried out periodically, for example every 6 seconds, every minute, every hour, etc.For example, at times ^4 and tn, the measured resistances are respectively equal to R2, R4 and Rn ohms, all greater than the value R^ However, at a time h, the measured resistance is equal to R3 ohms, less than Rh Similarly, the value R„ is, for example, less than the value R4. The sequence of measured resistances is therefore not an increasing sequence. Nevertheless, the average of the values ​​R2, R3, R4 and R,i is, for example, greater than R^. It is then agreed that the value of the resistance of the element tested increases. On the contrary, if the average of the values ​​R2, R3, R4 and Rn is, for example, less than R\, it is then agreed that the value of the re- . resistance of the tested element decreases.

[0067] According to one embodiment, as soon as the resistance value belongs to the threshold interval 106 at an instant and when it is measured as being increasing following one or more other measurements, it is estimated that the resistive element tested is an HRS element. On the contrary, if the resistance decreases, following one or more other measurements, it is estimated that the element is LRS.

[0068] Other ways of estimating whether the resistance increases or decreases are of course conceivable. For example, an average of the differences in the resistance values ​​between two consecutive measurements is calculated, and if the average is negative, it is estimated that the element is an LRS element, and if the average is positive, it is estimated that the element is HRS. In another example, the average of a first number of resistance values, measured for a first number of consecutive times, is compared to the average of a second number of resistance values, measured for a second number of consecutive times, subsequent to the first times.

[0069] [Fig.7] is a block diagram illustrating a circuit 700 according to an embodiment of the present description.

[0070] The circuit 700 comprises a resistive element 701 of a resistive non-volatile memory. The resistive element 701 is supplied by a current In of predefined intensity, via a transistor 702. In certain cases, the transistor 702 is used as a current source, in order to control the level of programming current flowing through the resistive element 701. In other cases, the transistor 702 is an access transistor used to activate or deactivate the programming current, and when it is activated, the intensity of the programming current is for example determined by a current source (not illustrated in FIG. 7), positioned elsewhere in series with the resistive element 701, for example at the end of the line of the matrix of the resistive non-volatile memory. For example, the element 701 was initially programmed to an LRS state or to an HRS state. As described in connection with Figures 1-5, the resistance of element 701 fluctuates over time.The current In is a current of predefined intensity. For example, the current In is applied to the resistive element 701 at an instant .

[0071] Alternatively, instead of applying a current of predefined intensity to the resistive element, a predefined voltage is applied to the resistive element.

[0072] The circuit 700 further comprises a circuit 704 (VOLT. READER) configured to measure a voltage representative of the resistance of the element 701. For example, the circuit 704 is further configured, for example via a voltage divider, to calculate the value of the resistance of the element 701 from the measured voltage. For example, the circuit 704 comprises an analog-to-digital converter (ADC) configured to convert the measured voltage value into a digital value, and a memory in which the digital value is stored, which is representative of the resistance value of the resistive element 701. For example, the circuit 704 comprises a comparator (not shown in the figure) and is further configured to compare the digital value with the values ​​Rthi and Rth2 defining the interval 106. For example, when the calculated resistance value does not belong to the threshold interval 106, the circuit 700 is for example configured to reapply the current or the voltage to the resistive element 701.

[0073] In one example, the circuit 700 further comprises, for example, a memory 706 (MEM), for example a shift register, configured to receive the digital voltage measurements obtained by the circuit 704. For example, the memory 706 is configured to store a first measurement in a location 708 (MEM1). For example, the first measurement is a voltage value corresponding to a resistance belonging to the threshold interval 106. The memory is further configured to store one or more other measurements made by the circuit 704 in a location 710 (MEM2). For example, the measurements stored in the location 702 do not all belong to the threshold interval 106.

[0074] In another example, the circuit 704 further comprises an amplifier, for example an operational transconductance amplifier (OTA) configured to provide a current based on the voltage measured at the output of the element 701. The memory 706 of the previous example is then replaced by at least two capacitors each connected to a switch and powered by the current generated by the OTA. The charge of the capacitors is then a representation of the measured resistances of the element 701.

[0075] The circuit 700 further comprises a comparison circuit 712 (COMPARISON CIRCUIT). For example, the circuit 712 is configured to determine whether the resistance of the element 701 increases or decreases, for example by applying the comparisons described in relation to FIGS. 6A and / or 6B.

[0076] In another example, the estimation of the state in which the element 701 has been programmed is carried out by a network of fully connected neurons. For example, the neural network comprises 4 layers. The input layer comprises, for example, a number 11 of neurons, n corresponding to the number of measurements carried out. For example, n is a value between 2 and 10. The following two layers comprise, for example, 8n and 2n neurons respectively. The output layer comprises 1 neuron and the output indicates whether the estimated state is HRS or LRS. For example, the output value is equal to the value 1 in the case where it is estimated that the state is HRS, in other words, that the resistance has a tendency to increase, and is equal to the value 0 in the case where it is estimated that the state is LRS, in other words, that the resistance has a tendency to decrease.

[0077] According to one embodiment, the circuit 712 is connected to a programming circuit 714 (REPROG.). The circuit 712 is for example configured to provide a signal to the circuit 714, the signal encoding whether the resistance tends to increase, or to decrease. In another example, the comparison circuit 712 is configured to program the state of a bit of the circuit 714. For example, the bit is programmed to state 1 if the resistance tends to increase, and to 0 if the resistance tends to decrease, or vice versa.

[0078] The circuit 714 is then configured to reprogram the state of the element 701 on the basis of the information provided by the circuit 712. The element 701 is then programmed, by the circuit 714, towards the HRS state if it has been estimated that the resistance has a tendency to increase, and towards the LRS state if it has been estimated that the resistance has a tendency to decrease.

[0079] According to another embodiment, the circuit 712 provides the information, for example in the form of a signal or by programming a bit, to a memory 716. For example, the memory 716 is configured to store, in association with an indication of the address of the element 701 in the non-volatile memory, whether the element is estimated to be HRS or LRS.

[0080] In one example, and although not illustrated in [Fig.7], the circuit 700 includes a selection circuit for selectively connecting each memory cell of the non-volatile memory to the circuit elements 704 to 712 and 714.

[0081] The following table groups together results obtained when the estimations were carried out by a fully connected neural network comprising 4 layers of, respectively, 10, 80, 20 and 1 neurons. Following a measurement belonging to the interval [10°800; 25°000] ohms, 9 other measurements were carried out at consecutive time intervals of 1 minute. In total 10 measurements were therefore carried out, the first belonging to the interval [10°800; 25°000] ohms, and were supplied to the neural network. The neural network is configured to predict, on the basis of the 10 measurements, whether the state of the tested element is HRS or LRS. As an example, for the experiment, 14,784 elements of the memory were elements programmed in the HRS state and 15,061 elements were elements programmed in the LRS state.From these elements, the training and validation of the neural network are carried out on a majority of the elements, for example on 75% of the elements, for example chosen randomly. The remaining elements, corresponding for example to 25% of elements, are for example used to carry out a test of the method, in other words to validate the model on data that were not seen during training. The following table presents the results obtained during the validation of the model, on a total of 7,461 elements, not used during the training and training of the network. A "support" column includes the number of elements, . having been tested, i.e. having entered the interval [10°800; 25°000] ohms. A column "Fl-score" includes the Fl-scores of the experiment. A column "Recall" includes the recall value of the experiment. A column "precision" includes the ratio of correctly predicted elements for each class.

[0082] [Tables 1] Accuracy recall Fl-score support HRS 0.96 % 0.86 0.91 3°698 LRS 0.88 % 0.96 0.92 3°763

[0083] Various embodiments and variants have been described. Those skilled in the art will understand that certain features of these various embodiments and variants could be combined, and other variants will occur to those skilled in the art. In particular, this is the case with regard to the estimation of the growth or decrease of the resistance value. Similarly, the number of measurements as well as the time interval between two measurements can vary. The choice of the terminals defining the threshold interval is made as a function of the volatile memory and the measurement circuit 704 and is left to those skilled in the art.

[0084] Finally, the practical implementation of the embodiments and variants described is within the reach of the person skilled in the art from the functional indications given above. In particular, this is the case with regard to the implementation of the elements 704, 706 and 716, which may be analog or digital.

Claims

Claims

1. Method for estimating the resistive state in which a cell of a non-volatile resistive memory is programmed, the method comprising: a) a first measurement, by a measurement circuit (704) connected to the memory cell, of a first value representative of the resistance (Re}T\) of the cell; b) at least a second measurement, by the measurement circuit, of at least a second value representative of the resistance ( ^1^2' Re{I\ Re^T4, Re^T„) of the same memory cell; c) the comparison, by a comparison circuit (712), of the first and at least one second values representative of the resistance;and d) based on the comparison, estimating the resistive state of the cell, the estimate being either an estimate of a high resistance state if the at least one second representative value exhibits an increase relative to the first representative value of the resistance, or an estimate of a low resistance state if the at least one second representative value exhibits a decrease relative to the first representative value.;

2. Method according to claim 1, further comprising: - before carrying out step a), a first application, at a first instant (li), of a current ( / „) of predefined intensity or of a predefined voltage to the cell of the resistive memory, the first measurement being carried out on the basis of the first application; and - following the carrying out of step a) and before carrying out step b), at least a second application, at at least a second instant ( ^2' ^3' *4' ^n), subsequent to the first instant, of the current ( / „) or of a predefined voltage to the resistive element, the at least one second measurement being carried out on the basis of the at least one second application.

3. Method according to claim 1 or 2, further comprising, following the estimation of the state: - reprogramming the cell towards the estimated state.

4. Method according to claim 1 or 2, further comprising, following the estimation of the state: - storing in a memory (716), in association with the cell, a value indicating the estimated state.

5. A method according to any one of claims 1 to 4, further comprising, following measurement of the first representative value: - comparing the first representative value with a reference interval (106); and - if the first representative value is outside the reference interval, resuming the method in a new implementation of the step

6. dj. Method according to any one of claims 1 to 5, further comprising, before the comparison: - converting, by a converter, the first and the at least one second representative values into a first and at least one second digital value; - storing, in association with the cell, the first and the at least one second digital value in a memory (706, 708, 710).

7. A method according to any one of claims 1 to 5, wherein the comparison is performed on the basis of voltage values stored by a first and at least one second capacitor configured to store the first and at least one second representative values respectively.

8. The method of any one of claims 1 to 7, wherein the estimated state is a high resistance state (HRS) if upon comparison, it is determined that the first representative value is less than one of the at least one second representative value and wherein the estimated state is a low resistance state (LRS) if upon comparison, it is determined that the first representative value is greater than one of the at least one second representative value.

9. The method of claim 1 to 7, wherein the estimated state is a high resistance state (HRS) if upon comparison it is determined that the first representative value is less than the average of the at least one second representative value.

10. The method of claim 7, wherein the at least one second representative value comprises a first second value and a second second value, the second second value being measured at a time later than the second time, and wherein the estimated state is a high resistance state (HRS) if upon comparison it is determined that the average of the differences between the first second and the second second representative values and between the first second and the first representative values is positive and in which the estimated state is a low resistance state (LRS) if upon comparison it is determined that the average of the differences between the first second and second representative values and between the first second and first representative values is negative.

11. Method according to any one of claims 1 to 10, in which the second instant (02) is distant from the first instant (fj) by a duration of at least 3 seconds, and for example at least one minute.

12. A method according to any one of claims 1 to 11, wherein the first and at least one second representative resistance values are voltage or current values representative of the resistance of the cell.

13. A circuit comprising a non-volatile resistive memory comprising a cell (701), programmed to one of a plurality of states, the circuit further comprising: - a measuring circuit (704) configured to measure a first value representative of the resistance (T^e^i) of the cell and at least one second representative value; - a comparison circuit (712) configured to compare the first representative value with the at least one second representative value to estimate the state of the cell, the estimate being an estimate of a high resistance state if the at least one second representative value has an increase relative to the first representative value of the resistance or of a low resistance state if the at least one second representative value has a decrease relative to the first representative value.

14. The circuit of claim 13, further comprising a current source configured to apply a current 0n) of predefined intensity, or a voltage source configured to apply a predefined voltage, to the cell at a first instant (¾ then at at least a second instant 02), subsequent to the first instant, the measurement of the first representative value being carried out on the basis of the application at the first instant and the measurement of at least one second representative value being carried out on the basis of the application at the at least second instant.

15. A circuit according to claim 13 or 14, wherein the non-volatile memory is a phase change memory, or an oxide-based resistive memory, or a programmable metallization cell.

Images