Embodiments disclosed herein pertain to memory systems and memory programming methods.
Memory devices are widely used in electronic devices, such as digital cameras and personal audio players, for storing digital data. Many different types of memory are available, each using a different fundamental technology for storing data, and the memory may be volatile or non-volatile memory. Resistive random-access memory (RRAM), conductive-bridge random-access memory (CBRAM) and flash are examples of non-volatile memory.
Referring to
In some instances, a memory cell may fail to place in the reset state following the application of an original reset pulse (e.g., time t=1) as determined by a respective subsequent verification operation. In such a situation, subsequent reset pulses may be applied to the memory cells having the same current as the original reset pulse until the memory cell places in the reset state.
At least some embodiments are directed towards improved memory systems and memory programming methods as described further below.
Referring to
Controller 12 controls operations of writing, reading and re-writing data of memory 16 as well as interfacing with other components or circuitry, such as sources of data to be stored within memory 16. Controller 12 may access and process commands with respect to memory 16 during operations of an associated device. Example commands instruct the generation of reset and set voltage potentials which are applied to memory 16 in one embodiment. The set and reset operations are used to write data to memory (i.e., program the memory) and are both referred to as write operations in one embodiment. Controller 12 may also control the application of read and verify pulses to memory 16 to read and verify stored data in one embodiment.
In one embodiment, controller 12 is configured to process data, control data access and storage, issue commands, and control other desired operations. Controller 12 may comprise processing circuitry configured to execute programming provided by appropriate computer-readable storage media (e.g., memory) in at least one embodiment. For example, the controller 12 may be implemented as one or more processor(s) and/or other structure configured to execute executable instructions including, for example, software and/or firmware instructions. Other example embodiments of controller 12 may include hardware logic, PGA, FPGA, ASIC, state machines, and/or other structures alone or in combination with one or more processor(s). These examples of controller 12 are for illustration and other configurations are possible.
Access circuitry 14 is coupled with controller 12 and memory 16 and is configured to implement addressing (selection of columns and rows of memory 16), writing, reading, verifying and re-writing operations with respect to memory cells of memory 16 in one embodiment. For example, access circuitry 14 may receive instructions from controller 12 to select a specific block, page, word or byte of the memory 16 as well as to implement writing, reading, verifying and re-writing with respect to a plurality of cells of the selected block, page, word or byte. As discussed below, the access circuitry 14 may apply electrical voltage potentials to the memory 16 to perform write, read and verification operations in one embodiment.
Memory 16 includes a plurality of memory cells configured to store data, conductors electrically connected with the memory cells, and perhaps additional circuitry, for example circuits of the access circuitry 14. At least some of the memory cells are individually capable of being programmed to a plurality of different memory states at a plurality of moments in time. Memory 16 is accessible to the user and/or associated device for storage of digital information. The memory cells may be configured as non-volatile cells in some implementations and may have different electrical resistances corresponding to different memory states. In one specific example implementation, memory 16 is implemented as conductive bridge random access memory (CBRAM) and the memory cells are conductive bridge memory cells.
Memory 16 may be implemented in different arrangements in different embodiments. For example, the memory 16 may be implemented within a memory device, such as a chip, a portion of the chip (e.g., tiles and/or sub-tiles discussed below) or other arrangements. The memory device may also include controller 12 and/or access circuitry 14 or portions thereof.
Referring to
The example memory cell 20 includes a first electrode 22, memory element 21 and second electrode 24, and the electrodes 22, 24 comprise electrically conductive material. The illustrated embodiment of memory element 21 includes an electrically conductive source member or layer 26 and a dielectric layer 28 intermediate the electrodes 22, 24. In one embodiment, the source layer 26 is a Cu+ source layer (e.g., CuTe), example materials of the dielectric layer 28 include AlOx, HfOx, and ZrOx, and the bottom electrode 24 is titanium nitride (TiN). Other embodiments are possible. Electrode 22 may be coupled with or part of a conductive common source line or plate.
The memory cell 20 shown in
More specifically, a set programming operation may be performed by the application of a voltage potential/bias to electrode 22 which is more positive than the voltage potential/bias applied to electrode 24. The application of these signals causes inducement of Cu ions into dielectric layer 28 and formation of one or more electrically conductive structures 29 (e.g., filaments) through dielectric layer 28 and between conductive source layer 26 and electrode 24. The formation of the structures 29 provides the memory cell 25 in a low resistance state. In one embodiment, the structures 29 comprise material (e.g., copper) from the source layer 26.
A memory cell 20 having the conductive structures 29 may be programmed in a reset operation to a high resistance state by the application of a voltage potential/bias to electrode 24 which is more positive than the voltage potential/bias applied to electrode 22. The application of these signals cause Cu ions to return into source layer 26 and dissolves any electrically conductive structures 29 within dielectric layer 28, thereby increasing the electrical resistance of the memory element 21 between the electrodes 22, 24 and providing the memory cell 20 in a high resistance state.
Memory cell 20 being may be repeatedly written between the high and low resistance arrangements at different moments in time to store different data values corresponding to the different memory (e.g., resistive) states. In one embodiment, a current is passed through the memory cell 22 and sense circuitry may measure the current to determine the resistance and memory state of the memory cell 20.
Referring to
Referring to
The depicted tile 40 includes a memory array 42 of a plurality of memory cells 20 which may be individually addressed by WL drivers 44 and Y-MUX circuitry 45. The tile 40 additionally includes an LIO controller 46, Vcommon driver 47, write driver 49 and a sense amplifier 50 in the illustrated embodiment. Tile 40 includes sixty-four of individual circuits 48, 49 and 50 to interface with a plurality of memory cells 20 of array 42 in parallel in one embodiment. LIO controller 46 provides interfacing of the sense amplifiers 50 of a given bank of the tile 40 to a databus (not shown) which is shared between multiple banks and also interfaces with an I/O block of the memory chip. Plate driver 47 drives the plate voltage to the various voltage values utilized for reading and writing. The write driver 49 drives the bitline voltage to the various voltage values utilized for writing. Sense amplifiers 50 sense the memory states of memory cells 20 during read and verification operations.
Referring to
In
Once a memory cell has been programmed to the set state (e.g., at time t=0), it may thereafter be reprogrammed to the reset state. The reset pulses applied to implement the programming operation to the reset state may have different electrical characteristics in some embodiments. In the example embodiment shown in
In one embodiment, and as discussed in additional detail below, it is desired to use a minimal current which is needed to change the programming from the set state to the reset state to extend endurance of the memory cells. Accordingly, the first pulse of waveform 60 corresponding to time t=1 is configured to generate the minimal current to attempt to change the programming to the reset state. As mentioned above, the pulses of waveform 60 are applied to the gate of access transistor 30 and bias the access transistor 60 to provide desired current to the memory element 21 in an attempt to program the memory element 21. The application of the pulse at time t=1 to the gate 32 results in the application of a signal having minimal current to the memory element 21.
However, the minimal current may not successfully place the memory cell 20 in the reset state as determined by a verification operation at time t=2. In one embodiment, the subsequent pulses of waveform 60 are configured to result in the application of respective signals of increasing current to the memory element 21 until a verification operation determines that the memory cell 20 properly placed in the reset state. Following verification of a proper reset placement, the programming operation to the reset state may be ceased with respect to the placed memory cell 20.
In one embodiment, the application of a plurality of pulses of waveforms 60 to gate 32 at a plurality of moments in time to generate a plurality of corresponding programming signals or pulses having increasing current which are applied to the memory cell 20 may be referred to as a single reset programming operation. In one embodiment, controller 12 is configured to implement the programming and verification operations including controlling the word line drivers 44 to increase the voltages of the word line signals applied to the access transistors 30 of memory cells 20 which failed to place in the reset state.
In one embodiment, the reset programming operation including applying the increasing programming currents is ceased with respect to memory cells 20 after the cells 20 have successfully placed into the reset state. Accordingly, in one embodiment, the memory cells 20 which have been properly placed into the reset state are isolated and no longer receive the reset program signals while others of the memory cells 20 which failed to place into the reset state may continue to receive the reset program signals of increasing current.
As mentioned above, the minimal current is used to program a memory cell to a reset state in one embodiment to extend endurance of the memory cells. However, if the minimal current is not successful in resetting a memory cell, the current can be increased or ramped a plurality of times until the memory cell is placed in the reset state.
CBRAM memory cells drift to higher reset current over time as a result of cycling and some cells may drift to a point where the memory system is incapable of providing sufficient current to program the cells from the set state to the reset state and the cells fail. The endurance of the memory array is limited by these reset (HRS) fails.
In one described embodiment, the current applied to the memory cell may be ramped or increased and which may successfully program the memory cells to the reset state which otherwise would have remained stuck in the set state. Furthermore, the use of a minimal current at the onset in accordance with one embodiment provides that the memory cells are programmed using minimal currents for successful programming which may slow the drifting of the memory cells to needing higher currents to be successfully reset inasmuch as the use of higher than necessary currents may increase the drifting of the memory cells to use of higher reset currents for proper programming. Accordingly, at least some of the embodiments slow the drifting of memory cells to higher reset current and provide increased current when needed in order to continue cycling while avoiding use of excess current which results in an extension in endurance of the memory array.
Although example embodiments are described with respect to CBRAM memory, the described may also be applied to other types of memory including other non-volatile resistive random-access memory (RRAM), for example, which rely upon atomic displacements for changing memory state.
Referring to
In
Referring to
The graphical representations of
Referring to
Referring to
As shown in
Referring to
Referring to
The use of the ramped word line signals according to one embodiment extends cell endurance with a positive read window budget beyond one million cycles while single pulse cycling fails 3 sigma at 300 k. The application of ramped signals according to one embodiment provides almost an extra decade of cycling as represented by line 142 compared with use of single pulses having fixed voltages for reset operations.
In some embodiments, a memory system comprises a memory array comprising a plurality of memory cells individually configured to have a plurality of different memory states, access circuitry configured to apply signals to the memory cells to program the memory cells to the different memory states, and a controller to configured to control the access circuitry to apply a first of the signals to one of the memory cells to program the one memory cell from a first memory state to a second memory state different than the first memory state, to determine that the one memory cell failed to place into the second memory state as a result of the application of the first signal, and to control the access circuitry to apply a second signal to the one memory cell to program the one memory cell from the first memory state to the second memory state as a result of the determination, wherein the first and second signals have a different electrical characteristic.
In some embodiments, a memory system comprises a memory array comprising a plurality of memory cells individually comprising a memory element configured to have different electrical resistances corresponding to a plurality of different memory states of the individual memory cell, and access circuitry coupled with the memory array and configured to apply a plurality of signals to the memory cells to program the memory cells into the different memory states, the access circuitry configured to apply one of the signals to one of the memory cells to change the electrical resistance of the one memory cell from one memory state to another memory state and to apply a plurality of the signals at a plurality of moments in time to another of the memory cells to change the electrical resistance of the another memory cell from the one memory state to the another memory state.
In some embodiments, a memory programming method comprises first applying a first signal to a memory cell to attempt to program the memory cell from a first memory state into a second memory state, determining that the memory cell failed to place in the second memory state as a result of the first applying, and after the determining, second applying a second signal to the memory cell to program the memory cell from the first memory state into the second memory state.
In some embodiments, a memory programming method comprises first applying a first signal to a memory cell to program the memory cell into a first memory state, the first applying forming an electrically conductive structure within a memory element of the memory cell providing the memory cell in a low resistance state corresponding to the first memory state, second applying a second signal to attempt to program the memory cell into a second memory state different than the first memory state, determining that the memory cell failed to place into the second memory state as a result of the second applying, and as a result of the determining, third applying a third signal to the memory cell to program the memory cell into the second memory state, the third applying removing the electrically conductive structure within the memory element providing the memory cell in a high resistance state corresponding to the second memory state.
In some embodiments, a memory programming method comprises identifying a plurality of memory cells of a memory array to be programmed into a first memory state, applying a plurality of first signals to the identified memory cells to attempt to program the identified memory cells into the first memory state, after the applying, determining that at least one of the identified memory cells failed to place into the first memory state, and as a result of the determining, applying a second signal to the at least one identified memory cell to program the at least one identified memory cell into the first memory state.
In compliance with the statute, the subject matter disclosed herein has been described in language more or less specific as to structural and methodical features. It is to be understood, however, that the claims are not limited to the specific features shown and described, since the means herein disclosed comprise example embodiments. The claims are thus to be afforded full scope as literally worded, and to be appropriately interpreted in accordance with the doctrine of equivalents.
This application is a continuation of and claims priority to U.S. patent application Ser. No. 15/050,248, filed Feb. 22, 2016, now U.S. Pat. No. 10,147,486, which is a continuation of and claims priority to U.S. patent application Ser. No. 14/151,729, filed Jan. 9, 2014, now U.S. Pat. No. 9,269,432, the teachings of which are incorporated herein by reference.
Number | Name | Date | Kind |
---|---|---|---|
5211434 | Lanava | May 1993 | A |
5268863 | Bader | Dec 1993 | A |
5602789 | Endoh | Feb 1997 | A |
5953255 | Lee | Sep 1999 | A |
6137716 | Wik | Oct 2000 | A |
6252795 | Hansen et al. | Jun 2001 | B1 |
6381670 | Lee et al. | Apr 2002 | B1 |
6392916 | Choi et al. | May 2002 | B1 |
6396738 | Tamada | May 2002 | B1 |
6496417 | Shiau et al. | Dec 2002 | B1 |
6529038 | Lambertson | Mar 2003 | B2 |
6714457 | Hsu | Mar 2004 | B1 |
6773842 | Liu | Aug 2004 | B2 |
6879525 | Van Brocklin et al. | Apr 2005 | B2 |
7006384 | Banks | Feb 2006 | B2 |
7028240 | Bautista et al. | Apr 2006 | B1 |
7042045 | Kang | May 2006 | B2 |
7130210 | Bathul et al. | Oct 2006 | B2 |
7200064 | Boerstler et al. | Apr 2007 | B1 |
7206225 | Wu | Apr 2007 | B2 |
7593259 | Kim | Sep 2009 | B2 |
7630245 | Kikuchi | Dec 2009 | B2 |
7679980 | Happ et al. | Mar 2010 | B2 |
7796421 | Mani | Sep 2010 | B2 |
7830718 | Sarin | Nov 2010 | B2 |
7944730 | Chen | May 2011 | B2 |
8154904 | Sekar et al. | Apr 2012 | B2 |
8213259 | Reed | Jul 2012 | B2 |
8271855 | Norman | Sep 2012 | B2 |
8279675 | Lee | Oct 2012 | B2 |
8331127 | Chen | Dec 2012 | B2 |
8392683 | Confalonieri | Mar 2013 | B1 |
8411503 | Lee | Apr 2013 | B2 |
8446776 | Hwang et al. | May 2013 | B2 |
8467226 | Bedeschi et al. | Jun 2013 | B2 |
8644060 | Abedifard et al. | Feb 2014 | B2 |
8830753 | Kim | Sep 2014 | B2 |
8830763 | Seo | Sep 2014 | B2 |
8836004 | Huang | Sep 2014 | B2 |
8850102 | Hanhimaki | Sep 2014 | B2 |
8854872 | Lam | Oct 2014 | B2 |
8879318 | Ahn | Nov 2014 | B2 |
8897075 | Choi | Nov 2014 | B2 |
8917550 | Lee | Dec 2014 | B2 |
8923056 | Kim | Dec 2014 | B2 |
8934298 | Park | Jan 2015 | B2 |
8934304 | Cho et al. | Jan 2015 | B2 |
8971088 | Jo et al. | Mar 2015 | B1 |
8995167 | Kim et al. | Mar 2015 | B1 |
9013934 | Lee | Apr 2015 | B2 |
9064580 | Senoo | Jun 2015 | B2 |
9123414 | Kitagawa et al. | Sep 2015 | B2 |
9183911 | Alam | Nov 2015 | B2 |
9208874 | Lee et al. | Dec 2015 | B2 |
9230685 | Strand et al. | Jan 2016 | B2 |
9269432 | Faraoni et al. | Feb 2016 | B2 |
9336875 | Fackenthal et al. | May 2016 | B2 |
9484097 | Kim | Nov 2016 | B2 |
9633728 | Kitagawa et al. | Apr 2017 | B2 |
9773551 | Strand et al. | Sep 2017 | B2 |
9837151 | Fackenthal et al. | Dec 2017 | B2 |
10121539 | Kitagawa et al. | Nov 2018 | B2 |
10147486 | Faraoni et al. | Dec 2018 | B2 |
10176868 | Kitagawa et al. | Jan 2019 | B2 |
10304531 | Strand et al. | May 2019 | B2 |
10311953 | Fackenthal et al. | Jun 2019 | B2 |
10770143 | Kitagawa et al. | Sep 2020 | B2 |
20050024938 | Ono | Feb 2005 | A1 |
20060245262 | Li | Nov 2006 | A1 |
20080080238 | Yuda | Apr 2008 | A1 |
20080165571 | Lung | Jul 2008 | A1 |
20080198673 | Fujita | Aug 2008 | A1 |
20090129140 | Kawazoe | May 2009 | A1 |
20100061166 | Tan | Mar 2010 | A1 |
20110001551 | Abou-Khalil | Jan 2011 | A1 |
20110235420 | Sharon | Sep 2011 | A1 |
20110267895 | Lee | Nov 2011 | A1 |
20120120711 | Rabkin et al. | May 2012 | A1 |
20120131400 | Rey-Losada | May 2012 | A1 |
20120230085 | Kawai et al. | Sep 2012 | A1 |
20120236624 | Costa et al. | Sep 2012 | A1 |
20120320656 | Chung | Dec 2012 | A1 |
20130021845 | Eleftheriou | Jan 2013 | A1 |
20130028021 | Sharon | Jan 2013 | A1 |
20130044535 | Shimakawa | Feb 2013 | A1 |
20130148409 | Chung | Jun 2013 | A1 |
20130163322 | Lam | Jun 2013 | A1 |
20130250657 | Haukness et al. | Jun 2013 | A1 |
20130201750 | Lee et al. | Aug 2013 | A1 |
20130223131 | Takagi | Aug 2013 | A1 |
20130250654 | Sugimae et al. | Sep 2013 | A1 |
20130286710 | Hall | Oct 2013 | A1 |
20130294165 | Park | Nov 2013 | A1 |
20130336069 | Kim | Dec 2013 | A1 |
20140003120 | Liao | Jan 2014 | A1 |
20140009997 | Toda | Jan 2014 | A1 |
20140078811 | Kawai | Mar 2014 | A1 |
20140112054 | Shimakawa et al. | Apr 2014 | A1 |
20140149824 | Ordentlich | May 2014 | A1 |
20150179256 | Kitagawa et al. | Jun 2015 | A1 |
20150194212 | Faraoni et al. | Jul 2015 | A1 |
20160172031 | Faraoni et al. | Jun 2016 | A1 |
20180197606 | Di Vincenzo | Jul 2018 | A1 |
20190279714 | Fackenthal et al. | Sep 2019 | A1 |
20200321043 | Em | Oct 2020 | A1 |
20200326880 | Akel | Oct 2020 | A1 |
20200327938 | Xue | Oct 2020 | A1 |
20200335163 | Jung | Oct 2020 | A1 |
Number | Date | Country | |
---|---|---|---|
20190074060 A1 | Mar 2019 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 15050248 | Feb 2016 | US |
Child | 16176417 | US | |
Parent | 14151729 | Jan 2014 | US |
Child | 15050248 | US |