Priority Information

fOOO J j This application claims priority to U.S. Application Number

13/746,768, filed January 22, 2013, which is a Non-Provisional Application of U.S. Provisional Application Number 61/734,473, filed December 7, 2012, the contents of which are herein incorporated by reference.

Technical Field

[0002] The present disclosure relates generally to semiconductor memory and methods, and more particularly, to apparatuses and methods related to stopping criteria for layered iterative error correction.

Background

|0ΘΘ3] Memory devices are typically provided as internal,

semiconductor, integrated circuits in computers or other electronic devices. There are many different types of memory including volatile and non- volatile memory. Volatile memory can require power to maintain its data (e.g., host data, error information, etc.) and includes random-access memory (RAM), dynamic random access memory (DRAM), and synchronous dynamic random access memory (SDRAM), among others. Non-volatile memory can provide persistent data by retaining stored data when not powered and can include NAND flash memory, NOR flash memory, read only memory (ROM), Electrically Erasable Programmable ROM (BEPROM), Erasable Programmable ROM (EPROM), and resistance variable memory such as phase change random access memory (PCRAM), resistive random access memory (RRAM), and magiietoresistive random access memory (MR AM), among others.

(0004j Memory devices can be combined together to form a storage volume of a memory system such as a solid state dri ve (SSD). A solid state drive can include non-volatile memory (e.g., . AND flash memory and. NOR. flash memory), and/or can include volatile memory (e.g., DRAM and SRAM), among various other types of non-volatile and volatile memory.

|0005{ An SSD can be used to replace hard disk dri ves as the main storage volume for a computer, as the solid state drive can have advantages over hard drives in terms of performance, size, weight, ruggedness, operati ng temperature range, and power consumption. For example. SSDs can ha e superior performance when compared to magnetic disk drives due to their lack of moving parts, which may avoid seek time, latency, and other electromechanical delays associated with magnetic disk drives.

10066] Memory is utilized as volatile and non-volatile data storage for a wide range of electronic applications. Non-volatile memory may be used in portable electronic devices, such as laptop computers, portable memory sticks, digital cameras, cellular telephones, portable music players such as MP3 players, movie players, and other electronic devices. Memory cells can be arranged into arrays, with the arrays being used in memory devices,

[0007] One type of error correction involves low density parity check

(IJD.PC) codes. Uueneoded (e.g., "raw") data can. be encoded into codewords for transmission and/or storage. The codewords can subsequently be decoded to recover the data. Powerful error correction may be desired but balanced against latency, throughput, and/or power constraints such as those imposed by portable electronic devices.

Brief Desc iption of the Drawings

f OO Sj Figure .1 is a block diagram of an apparatus in the form of a computing system including at least one memory system in accordance with a number of embodiments of the present disclosure,

[00Θ9] Figure 2 illustrates a flow diagram for stopping layered iterative error correction in accordance with a number of embodiments of the present disclosure.

[0010! Figure 3 is a lot illustrating error correction circuitry (ECC) power savings versus raw bi error rate (RBER) according to a number of embodiments of the present disclosure.

[0011] Figure 4 is a plot illustrating codeword error rate (CWER.) versus ra bit error rate (RBER) according to a number of embodimen ts of the present disclosure. Detailed Description

[0Q12] The presen disclosure includes apparatuses -and methods related to stopping criteria for layered iterative error correction. A number of methods can include receiving a codeword with an error correction circuit, iteratively error correcting the codeword with the error correction circuit including parity checking the codeword on a layer-by-layer basis and updating the codeword after each layer. Methods can include stopping the iterative error correction in response to a parity check being correct for a particular layer,

[0013] In the following detailed description of the present disclosure, reference is made to the accompanying drawings that form a part hereof, and in which is shown by way of illustration how one or more embodiments of the disclosure may be practiced. These embodiments are described in sufficient detail to enable those of ordinary skill in the art to practice the embodiments of this disclosure, and it is io be understood thai other embodiments may be utilized and that process, electrical,, and/or structural changes may be made without departing from the scope of the present disclosure. As used herein, the

designator "M", particularly with respect to reference numerals in the drawings, indicates that a number of the particular feature so designated can be included. As used herein, "a number οΓ a particular thing can refer to one or more of such things (e.g., a number of memory devices can refer to one or more memory devices).

[0014) The figures herein follow a numbering convention in which the first digit or digits correspond to the drawing figure number and the remaining digits identify an element or componen in the drawing. Similar elements or components between different figures may be identified by the use of similar digits. As will be appreciated, elements shown in the various embodiments herein can be added, exchanged, and/or eliminated so as to provide a number of additional embodiments of the present disclosure. n addition, as will be appreciaied, the proportion and the relative scale of the elements provided in the figures are intended to illustrate certain, embodiments of the present invention, and should not be taken in a limiting sense.

|UU15j Figure 1 is a block diagram of an apparatus in the form of a computing system 100 including at least one memory system 104 In accordance with a number of embodiments of the present disclosure. As used herein, a memory system 104, a controller 108, or a memory device 1. 10 might also be separately considered an "apparatus." The memory system 104 can be a solid state drive (SSD), for instance, and can include a host interface 106, a controller 108 (e.g., a processor and/or other control circuitry), and a number of memory devices 1 .10-1 , . . ., .1 10-M (e.g., solid state memory devices such as NA D Flash devices), which provide a storage volume for the memory system 104. The number of memory devices i 10-1 , . . ,, 1 10-M may be referred to generally herein as memory device 1 10. m another embodiment, the memory system 104 may be a. single memory device.

[0016] As illustrated in Figure 1, the controller 108 can be coupled to the host imerface 106 and to the memory device 1 10 via a plurality of channels and can be used to send data between the memory system 1 4 and a host 102. The interface .106 can be in the form of a standardized interlace. For example, when the memory system 3 4 is used for data storage in a computing system 100, the interface 106 can be a serial advanced technology attachment (SAT A), peripheral component interconnect express (PCle), or a universal serial bits (USB), among other connectors and interfaces. In general, however, interface 106 can provide an interface for passing control, address, data, and other signals between the memor system 104 and a host 102 having compatible recepiors for the interface 106.

(00! 7j The host 102 can be a host system such as a personal laptop computer, a desktop computer, a digital camera, a mobile telephone, or a memory card reader, among various other types of hosts. The host 102 can include a system motherboard and/or backplane and can include a number of memory access devices (e.g., a number of processors). The host 102 can also be a memory controller, such as where memory system 104 is a memory device (e.g., having an o.u-die controller).

(0018| The controller 108 can communicate with the memory device ί 10

(which in some embodiments can be a number of memory arrays on a single die) to control data read, program verily, write, and erase operations, among other operations, in some embodiments, the controller 108 can be cm the same die or a different die than any or ail of the number of memor devices 110-1 , . .„ 110-M. 10019] Although not specifically illustrated, in some embodiments, the controller 108 can include a discrete memory channel controller for each channel coupling the controller 108 to the memory devices i.1.0-1. . , ., 1. 1.0~M. The controller 108 can include, for example, a number of components in the form of hardware and/or firmware (e.g., one or more integrated circuits) and/or software for controlling access to the number of memory devices 1 10- L 1 10- and/or for facilitating data transfer between the host 102 and memory devices 1 10- ! , 110-M.

(0Θ2Θ) The number of memory devices 1 10- 1 , . . ., i 10-M can include a number of arrays of memory cells (e.g., non-volatile memory ceils). The arrays can be Flash arrays with a HAND architecture, for example. However, embodiments are not. limited to a particular type of memory array or array architecture. The memory cells can be grouped, for instance, into a number of blocks including a number of physical, pages. A number of blocks can. be included in a plane of memory cells and an array can include a number of planes. As one example, a memory device may be configured to store 8KB (kilobytes) of user data per page, 128 pages of user data per block, 2048 blocks per plane, and 1 6 planes per device.

[002ij In operation, data can be written to and/or read from memory

(e.g., memory devices 1 10-1 , . . ., U 0-M of system 104) as a page of data, for example. As such, a page of data can be referred to as a data transfer size of the memory system. Data can be transmitted to/from a host (e.g., host 102) in data segments referred to as sectors (e.g., host sectors). As such, a sector of data can be referred to as a data transfer size of the host.

(0022] As illustrated in Figure 1, the controller 1.08 can include an.

iterative error correction circuit 1 12 ("Layered iterative ECC") coupled to an. algebraic error correction circuit i 14 ("Algebraic ECC). The iterative error correction circuit 1 12 can be coupled to the memory device 1 30 and the algebraic error correction, circuit 1 14 can be coupled to the host interface 106. For example, the iterative error correction circuit 1 12 can be a quasi-cyclic low density parity check. (LDPC) code circui , among other iterative error correction, circuits thai apply layered iterative error correction. Although illustrated as a single circuit, the iterative error correction circuit 1 12 can be provided as a separate encoding circuit (e.g., for encoding data to be stored in the memory device 1 10) and decoding circuit (e.g., for decoding data stored in the memory device 1 10). The algebraic error correction circuit 1 14 can be one of the group including a B<¾e-Chaudh¾ri-Hocqueng em (BCH) error correction circuit, and a Reed Solomon error correction circuit, among other algebra c error correction, circuits. Although illustrated as a single circuit, Che algebraic error correction circuit 1 14 can be provided as a separate encoding circuit (e.g., for encodin data to be stored in the memor device 1 10) id decoding circuit (e.g., for decoding data stored in the .memory device 110). Each of the iterative error correction circuit 1 1 and the algebraic error correction circuit 1 1 can be discrete components such as an application specific integrated circuit (ASIC) or the components may reflect functionally provided by circuitry within the controller 108 that does not. necessarily have a discrete physical form separate from other portions of the controller 108. Although illustrated as components within the controller 108 in Figure L each of the i terative error correction circuit 1 12 and the algebraic error correction circuit i 14 can be external to the controller 108 or have a number of components located within the controller 108 and a number of components located external to the controller 108.

[00233 The iterative error correction circuit i 12 can be configured to receive a codeword from the memory device 1 1 . in a number of embodiments, the iterative error correction circuit 1 12 can receive the codeword, which can be compri sed of hard data read from the memory device 110, without receiving soft data, in a number of embodiments, the iterative error correction circuit 1 12 can receive the codeword comprising hard data and can receive soft data

corresponding to the codeword therewith. In a number of embodiments, the iterative error correction circuit 1 12 can receive the codeword comprising hard data and can. separately receive soft data corresponding to the codeword (e.g., where the soft da ta is received based on a separate soft read of the codeword or where the soft data was originally read with the hard data, buffered, and received ^" from the buffer). Hard data is data that corresponds to the data state of a memory cell. For example, a 2-bit memory cell can be programmed to one of four data stales, where each data state corresponds to one of data 00, 01 , 10. or 1.1. In contrast, soft data associated with a memory cell can indicate a location of a state (e.g., threshold voltage (Vt)) stored on the memory cell within a distribution of states (e.g., Vt distribution) representing the target, state to which the memory cell was programmed. Soft data corresponding to a codeword ( where the codeword is comprised of hard data) can include soft data that corresponds to (he hard data, in ihe codeword (e.g., the soft data can indicate a greater resolution of the hard data in the codeword). A memory device can be configured to determine a particular number of soil data bits for hard data read therefrom. For the 2-bit memory cell described above, an example of soft data could include a greater resolution thai uses four bits, which could otherwise represent up to sixteen different states. The resolution of the soft data can depend on the number of reads made to iocate the Vt stored on the memory ceil for those embodiments employing discrete read s gnals, where more reads can provide greater .resolution. For those embodiments employing a ramping read signal, the resolution of the soft data can depend on a resolution provided as an output from an analog-to-digital converter or other circuitry thai con-elates a value of the Vl of the memory cell to data, where the resolution is reflected by the number of bits transmitted across an input/output (I/O) bus from the memory device S 10, For example, a memory device 1 10 can be configured to read a particular memory cell that stores two bits of data with a resolution of eight bits and to transfer anywhere from, two bits (two bits of hard data and zero bits of soft data) to eight bits (two bits of hard data and six. bits of soft data.) across an I/O bus,

|0024| The iterative error correction circuit 1 12 can iteratively error correct the codeword, by parity checking the codeword on a layer-by-layer basis and updating the codeword after each layer. One iteration of error correction performed by the iterative error correction circuit 1 .12 can include more than one parity checking layer. The iterative error correction circuit 112 can use an iterative belief propagation decoder, where the iterative error correction circuit 1 12 can include two types of processing nodes: variable nodes and check nodes. The processing nodes can be coupled according to the LDPC code's parity check matrix. The check nodes can be implemented as circuits that parity check (e.g., perform a syndrome check) and/or update the codeword. In a number of embodiments, updating the codeword can include updating reliability data (e.g., a log-likelihood ratio (LL.R)), among other functions. For example, updating the reliability data associated with each variable node can cause a value of respective bit of the codeword that is associated with each variable node to be updated (e.g., where updating a bit of the codeword includes changing a value of the bit or holding the value of the bit). Thus, updating the reliability data effectively updates the codeword For embodiments that include receiving hard date without, soft data and/or reliability data (e.g., based on hard data reads), reliability data cart be assigned to the codeword based, at least in part, on the LDPC code's parity check constraints during the decoding process. For embodiments thai include receiving bard data with soft data (e.g., based on soft data reads), the reliability data can be assigned to the codeword based, at least in part, on the soft data (e.g., the input to the iterative error correction circuit 1 12 will include reliability data, unlike the hard read case). The. codeword can be updated during each layer (or some multiple of the layers, or once per iteration, for example) of decoding (e.g., the iterative error correction circuit 1 12 can update the input reliability data based on the parity check constraints of the LDPC code). Each layer of an iteration can include recei v ing an input from less than all (e.g., a respective one) of the check nodes and updating the reliability data based, at least in part (e.g., in some embodiments, based only), on the input (e.g., updated reliability data) from the respecti ve one of the check nodes. Some embodiments can include updating the reliability data every L-itumber (e.g., where L. is a variable from 1 layer to ail layers, where the reliability data can be updated every layer, every other layer, every third layer, and so on, up to once per iteration) of the number of layers. The reliability data can be updated based on the input .from the respective ones of the check nodes coupled thereto, hi a number of embodiments, the layered updating can proceed until the variable node has received an input from each of the cheek nodes coupled thereto. The layered updating of the stored reliability data value at the variable node for ail. of the number of check nodes coupled thereto can be completed for one iteration, before sending a subsequent (updated) reliability data value from the variable node to each of the check nodes coupled thereto.

(0025] The iterative error correction circuit. 1 12 can transfer the codeword or a copy thereof to the algebraic error correction c ircuit 1 1 in response to the layered parity check being correct for a particular layer. In some embodiments, the iterative error correction circuit. 1.12 can. sto error correcting the codeword in response to the layered parity check being correct for a particular layer or in response to the layered parity check being correct for the particular layer and at least one other layer. In a number of embodiments, the layered parity check can be considered to be correct when there are no parity errors to the particular layer, of in some embodiments, in the particular layer and at. least one other layer, (where the parity check is correct for all. units of data checked in. the particular layer and/or at least one other layer). In some embodiments, the layered parity check is correct when there are less than a. threshold number of pari ty errors in the particular layer, or in some

embodiments, in the particular layer and at least one other layer. For example, the threshold number of parity errors can. be selected based on the correction capability of the algebraic error correction circuit 1 14 In a number of

embodiments, the controller .108 can be configured to select the threshold number of errors based on a number of characteristics of the memor device 1 10, such as a temporal age of the memor ' device 1 10, a number of

program erase cycles of the memory device 1 .10, a storage density of the memory device 1 1 . a retention rate of (he memory device 1 1 (e.g., how long programmed memory ceils can retain a charge), and/or a physical location within the memory device .1 .1 where the codeword is stored (e.g., whether the codeword , or a, portion thereof is stored on an edge word line), among other characteristics of the memor device 1 1 . in some embodiments, the threshold number of parity errors can be defined in firmware of the controller I0S as a selectable option,

[0026] in some embodiments, the iterative error correction circuit 1 12 can error correct the codeword until the codeword includes less than a threshold number of parity errors in a particular layer of a particular iteration. According to a number of embodiments, a concatenated coding scheme can be employed such that the algebraic error correction circuit 1 14 can receive the updated codeword from the iterative error correction circuit 112 (e.g., after the codeword includes less than a threshold number of parity errors in a particular layer). Then the algebraic error correction circuit 1 14 can. error correc the codeword according to its algebraic error correction code constraints. The algebraic error correction circuit 1 14 can be configured, to transfer the codeword or a copy thereof to the host interlace 106 (and the host interface can be configured to receive the codeword) in response to a successful error correction by the algebraic error correction circuit.. The controller 108 can be configured to indicate a successful program verify operation in response to a successful error correction by the algebraic error correction circuit. In the event that the error correction by the algebraic error correction circuit 114 is -unsuccessful, the iterative error correction circuit 1 12 can be configured to receive soil data from the memory device 1 JO {e.g., from a soft read of the memory or from a buffer that stores previously read soft data) in response to the unsuccessfid error correction by the algebraic error correction circuit 1 14 (e.g.. the controller 108 can initiate a soft read and/or request soft data that was previously read along with the hard data).

|0027 j In a number of embodiments, the iterative error correction circuit

1 12 can be iterated only once for a received codeword comprising hard data and then the codeword can be transferred to the algebraic error correction circuit 1 14 for any remaining error correction not accomplished by the iterative error correction circuit 1 12. if the concatenated scheme is insufficient to correct the codeword, soft data can be retrieved from the memory device 1 1 and the iterative error correction circuit 1 12 can be iterated as many times as necessary to correct the codeword or, up to a predetermined threshold number of iterations.

[00283 The iterative error correction circuit i 12 can deliver increased performance over a number of iterations (e.g., iterations of a normalized-min- sum or other iterative decodin algorithm). An iteration of an. iterative error correction circuit i 12 can be equivalent to going over all parities and updating the codeword (e.g., hard decisions). However, the amount of power consumed by tiie iterative error correction circuit 1 12 is directly proportional to the number of iterations performed. Given the stringent power budget of some memory systems (e.g., systems employing flash memory such as in portable electronic devices), number of embodiments of the present disclosure stop the iterative error correction circuit 1 12 prior to completing ail iterations, or even prior to completing a particular iteration. Furthermore, a concatenated coding scheme is employed that uses both the iterative error correction circuit 112 (e.g., providing an inner code) and the algebraic error correction circuit 1 1 (e.g., providing an outer code). Given that the iterative error correction circuit 1 12 and the algebraic error correction circuit 1 14 both work to correct a codeword, stopping the iterative error correction circuit 112 as soon as the remaining errors are suspected to be within the correction capability of the algebraic error correction circuit 11.4 can be beneficial in terms of power savings. Error correction of the codeword can begin with the algebraic error correction circuit in response to stopping iterative error correction with the iterative error correction circuit (e.g.. when a particular layer has less than a threshold number of parity errors).

However, in number of embodiments, the iierati ve error correction circuit can. continue to iteratively error correct the codeword while the algebraic error correction circuit operates on (e.g., corrects) the codeword. In such

embodiments, the iierative error correction circuit can be stopped i response to the second error correction circuit reporting an uncorrec able error for the codeword. By way of example, the iierative error correction circuit 1 .12 cat) comprise an LDPC decoder thai employs a layered min-sum decoding algorithm, where a layer can. be considered a fraction of an iteration at which each LDPC codeword bit is updated based on a single parity check, if such a layer results in no parity errors (or a less than a threshold number of parity errors), an assumption can be made that the codeword will be within an error correction capability of the algebraic error correction circuit i 14.

[0029] Figure 2 illustrates a flow diagram for stopping layered iterative error correction in accordance with a number of embodiments of the present, disclosure. At 220, a codeword comprising hard data can be received from a memory device (e.g., memory device 1 10 illustrated in Figure I). At 222, error correction can be performed on the codeword (e.g., with an iterative error correction circuit 1 12 illustrated in Figure 1). For example, iterative error correction can be performed on the codeword on a layer-by-layer basis. After layer "L", the codeword can be updated as illustrated at 224. After each layer of error correction, the iterative error correction circuit can be aware of the number of parity errors In. that layer. At 226, if the number of parity errors is within a parity check threshold (e.g., selected based on the correction capability of an algebraic error correction circuit), then the iterative error correction can be stopped (without checking a next layer of the iteration) and the codeword can be transferred to a different en-or correction circuit (to the algebraic error correction circuit) at 230, At 226, if the number of parity errors is not within a parity check threshold, then a next layer of the error correction iteratio can be performed as illustrated at 228 and 222. Although not specifically illustrated in Figure 2, the iterative error correction can be stopped in response to a threshold number of i terations being completed (regardless of whether the parity check is correct for a last layer of a last iteration of error correction).

I S (003Θ] At 232, after the algebraic error correction circuit has error corrected the codeword, a determination can be made as to whether the codeword passed or failed the algebraic error correction. If the algebraic error correction failed, at 234, soft data corresponding to the codeword can be retrieved iron ^' ! the memory device arid iterati ve error correction can begin for the codeword taking into account the soft data at 222. in a number of embodiments of the present disclosure, soft data can be recei ved from the memory device in response to failed error corrections by the first (e.g., the iterative error correction circuit) and the second (e.g., the algebraic error correction circuit) error correction circuits. The algebraic error correction circuit can correct up to a threshold number of errors (e.g., bit or symbol errors). Although not specifically illustrated in Figure 2, the algebraic error correction circuit can report art uncorrectable error in response to a number of errors in t he codeword exceedi n g a threshold number of errors correctable by the algebraic error correction circuit. If the algebraic error correction succeeded, at 236, the codeword can be output to the host. The codeword can be transferred to the host in response to successful error corrections by the first and the second error correction circuits.

10031 Figure 3 is a plot illustrating error correction circuitry (ECC) power savings versus raw bit error rate (RBER) according to a number of embodiments of the present disclosure. The ECC power savings is represented from 0 to P where P is indicative of a percentage of power savings (not necessarily 100%). The RBER is represented irom 10 ^*(K* ° to 1 ^l where x represents a positive integer. The RBER closer to the right of the graph can be representati ve of a typical RBER at the beginning of life for a typical memory device. As is illustrated, the power savings can diminish with increased raw bit error rate (trending toward the left side of the graph). The curve represents experimental data indicating the power savings for an iterative error correction circuit according to a number of embodiments of the present disclosure.

[0032] Figure 4 is a plot illustrating codeword error rate (CWER) versus ra bit error rate (RBER) according to a number of embodimen ts of the present disclosure. The CWER is represented from 10" to Mf. The RBER is represented from (y l)*^* * to (y-l)10 ^{"ix": i>} , where x and y each represent positive integers. The value of x in Figure 4 is not necessarily the same as the value of x in Figure 3. Although not distinguishable in Figure 4, two curves are actually shown. A first: curve represents the codeword failure rate when the iterative error correction circuit is stopped according to a number of

embodiments of the present disclosure and a second curve (essentially collinear with the .first curve) represents the codeword failure rate according to some previous approaches that do not stop the iterative error correction circuit, prior to the completion of a number of iterations according to known algorithms (e.g., the iterative error correction circuit is not stopped early). Thus, as illustrated in Figure 3 and Figure 4, a number of embodiments of the present disclosure can decrease the power consumption of an iterative error correction circuit (e.g., in a concatenated error correction scheme) without increasing the codeword failure rate.

Conclusion

10033] The present disclosure includes apparatuses and methods related to stopping criteria for layered iterative error correction. A number of methods can include receiving a codeword with an error correction circuit, iferatively error correcting the codeword with the error correction circuit including parity checking the codeword on a layer-by -layer basis and updating the codeword after each layer. Methods can include stopping the iterative error correction in response to a parity check being correct for a particular layer.

[0034] Although specific embodiments have been illustrated and described herein, those of ordinary skill in the art will appreciate that an arrangement, calculated to achieve the same results can be substituted for the specific embodiments shown. This disclosure is intended to cover adaptations or variations of one or more embodiments of the present disclosure, it is to be understood that the above description has been made in an illustrative fashion, and not a restricti ve one. Combination of the above embodiments, and other embodiments not specifically described, herein will be apparent to those of skill in the art upon, reviewing the abo ve description. The scope of the one or more embodiments of the present disclosure includes other applications in which the above structures and methods are used. Therefore, the scope of one or more embodiments of the present disclosure should be determined with reference to (he appended claims, along with the full range of equivalents to which such claims are entitled

[0035] In the foregoing Detailed Description, some features are grouped together in a single embodiment for the purpose of streamlining the disclosure. This method of disclosure is not to be interpreted as reflecting an intention that the disclosed embodiments of the present disciosure have to use more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive subject matter lies in less than all features of a single disclosed embodiment. Thus, the following claims are hereby incorporated into the Detailed Description, with, each claim standing on its own as a separate embodiment.