|Publication number||US20080181030 A1|
|Application number||US 11/862,412|
|Publication date||Jul 31, 2008|
|Filing date||Sep 27, 2007|
|Priority date||Jan 31, 2007|
|Also published as||CN101236516A, US8045405, US20080184002|
|Publication number||11862412, 862412, US 2008/0181030 A1, US 2008/181030 A1, US 20080181030 A1, US 20080181030A1, US 2008181030 A1, US 2008181030A1, US-A1-20080181030, US-A1-2008181030, US2008/0181030A1, US2008/181030A1, US20080181030 A1, US20080181030A1, US2008181030 A1, US2008181030A1|
|Original Assignee||Samsung Electronics Co., Ltd.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Referenced by (1), Classifications (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This is a divisional of application Ser. No. 11/779,349 filed on Jul. 18, 2007, which is incorporated herein by reference in its entirety.
1. Field of the Invention
The invention relates generally to memory systems, memories, and command protocols for memory systems.
2. Description of Related Art
The concept of a “memory system” now encompasses a great variety of circuits and related control methods and protocols enabling the transfer, storage and retrieval of digital data. Once memory systems were associated with only computers systems and similar computational logic platforms. Now, a great host of consumer products ranging from cell phones to automobiles to refrigerators include memory systems of varying degrees of complexity.
A generic memory system is conceptually illustrated in Figure (FIG.) 1, wherein a memory 2 stores data received from a memory controller 1 via a channel 3.
As defined by its application, memory 2 may be volatile or nonvolatile in its operating nature. A volatile memory retains stored data only so long as power is applied to the memory. Dynamic Random Access Memories (DRAMs) and Static Random Access Memories (SRAMs) are well known examples of volatile memories. In contrast, nonvolatile memories have the ability to retain stored data in the absence of applied power. NAND and NOR type flash memories are examples of nonvolatile memories.
Regardless of memory type and related storage capabilities, a memory must typically be associated with some kind of memory controller. Memory controller 1 shown in
Data transfer between memory controller 1 and memory 2 is accomplished via a channel 3. Channel 3 may be hardwired or wireless in its implementation. For example, data may be wirelessly transferred between memory controller 1 and memory 2 via radio frequency (RF) channel(s), infrared channel(s) and/or electromagnetic channel(s). More typically, memory controller 1 and memory 2 are connected via a hardwired channel formed by one or more buses and/or various signal lines. In this context, a “bus” is merely a collection (physical or operational) of signal lines commonly operated in relation to a block of data and/or a data transfer period.
The examples shown in
In the second illustrated example of
In the third illustrated example of
Finally, in the fourth illustrated example of
Regardless of the specific nature of channel 3, the transfer of data between memory controller 1 and memory 2 is made in relation to a defined protocol. A “protocol” is a system of rules that defines how something is to be accomplished. Thus, is the context of a memory system, a protocol is an agreed-upon or standardized “language” for communicating data and/or establishing a communications connection between a memory and some other device, such as a memory controller. Many memory system protocols are implemented, in whole or it part, according to industry standards. Other protocols are custom in their definition and use. In some instances, one protocol may be used to call (or invoke) another protocol.
One notable characteristic of conventional and emerging memory systems is an increasing demand for greater data bandwidth (available data per unit operation) and/or data throughput (available data per unit of time). Data bandwidth may be increased by increasing the number of data bits communicated to/from a memory per unit memory system operation (e.g., read and write operations). Data throughput may be increased by increasing the number of data bits communicated per unit memory system operation, and/or by increasing the speed with which each memory system operation is performed.
Given the concurrent commercial motivations of reducing memory system size and power consumption while maximizing available data bandwidth and data throughput, it is not surprising that certain practical limitations and resource conflicts have been identified. For example, the number of signal lines connecting a memory controller with a memory may be limited by the size of the memory or memory controller, and/or the corresponding number of available connection (input/output) pads. Given such physical limitations, all or some of the signal lines forming a channel in a memory system may be multiplexed in their use.
Regardless of the physical connections between a memory and memory controller and the controlling communications protocol, almost all contemporary memory systems are being run at increasingly fast clock speeds. Increasingly fast clock speeds facilitate greater data throughput, and enhanced data throughput is a highly desirable characteristic in many commercial applications.
Unfortunately, increasingly fast clock speeds also increase the likelihood of data communication (transmission and/or reception) errors. Indeed, system complexity has become so great and data transfer speeds so fast, that many contemporary memory systems now incorporate error detection and/or error correction (hereafter, singularly or collectively indicated as “EDC”) capabilities designed to mitigate the inevitable consequences of data communication errors.
EDC capabilities were once used primarily for long-haul (or bulk) data traffic such as telephone networks and satellite communication systems. Now, however, the advantages of incorporating EDC capabilities within a memory system are well appreciated. There are many different types of EDC protocols, techniques, schemes, and related circuits and software. One class of relatively simple error detection techniques are conventionally used to implement a functionality referred to as a cyclic redundancy checker (CRC). More sophisticated EDC techniques are capable of not only detecting the presence of one or more error(s) in communicated data, but also of correcting the detected error(s).
Nearly all EDC techniques are implemented by adding additional (“overhead”) data bits to a block of data being communicated. That is, the data block is first run through a mathematical or logical calculation in order to produce corresponding EDC data. The EDC data is then transferred along with the data block. At the receiving end, the data block is again run through a similar mathematical/logical calculation and the resulting data compared to the received EDC data. A successful comparison indicates an error free block of data. A failed comparison indicates one or more errors in the data block. When more sophisticated EDC techniques are used, these error(s) may be corrected by further resort to the EDC data.
EDC capabilities are particularly beneficial in the context of memory systems transferring data in packet form. That is, individual data packets may be defined to include corresponding EDC data along with other types of data (e.g., control data, address data, write data, etc.). The EDC data may be associated with (e.g., derived from and used to detect and/or correct errors in) any one or more of these other data types contained in the data packet.
While EDC capabilities offer great benefits in the verification of data being communicated between a memory controller and a memory, such capabilities come at a price. The greatest price is typically imposed on the overall operating speed of the memory system. EDC operations run in the memory controller, and more particularly, EDC operations run in the memory potentially generate a data throughput bottleneck in the memory system. Thus, the contemporary memory system designer faces the competing demands of (1) increasing data bandwidth/throughput which require the streamlining, simplifying, and expediting of memory system operations, and (2) assuring data integrity which requires sophisticated and relatively slow EDC operations.
In recognition of the competing demands placed upon memory system designers, embodiments of the invention provide a memory system command protocol that offers improved immunity to communication errors. This is particularly true in the context of “write commands.” In one aspect, the improved immunity of memory system commands within embodiments of the invention allows a reduction in the sophistication of EDC capabilities incorporated within the memory system, or the completion omission of EDC capabilities from the memory system. Alternately, given a defined set of EDC capabilities within a memory system, a command protocol consistent with embodiments of the invention results in a significant reduction in the likelihood of command misinterpretation and erroneous execution in an associated memory. As a further result, data integrity is assured and memory system reliability improved.
In one embodiment, the invention provides a memory system comprising; a memory controller communicating a command to a memory, the command being selected from a set of commands comprising a write command and a plurality of non-write commands, wherein a Hamming distance value calculated between a digital value indicating the write command and a digital value indicating any one of the plurality of non-write commands is greater than 1.
In a related embodiment, each command in the set of commands may be uniquely indicated by command data and the command data indicating the write command may comprise a plurality of write enable bits.
In another embodiment, the invention provides a memory system comprising; a memory controller connected to a plurality of memory elements arranged on a memory card via a plurality of point-to-point interfaces, the memory controller communicating a command to at least one memory associated with at least one of the plurality of memory elements, the command being selected from a set of commands comprising a write command and a plurality of non-write commands, wherein a Hamming distance value between the write command and each one of the plurality of non-write commands is greater than 1.
In a related embodiment, the plurality of memory elements may comprise a stacked arrangement of memories including the memory, wherein a lowermost memory in the stacked arrangement re-drives the command to at least one other memory in the stacked arrangement of memories.
In another embodiment, the invention provides a method of controlling operation of a memory, the method comprising; selecting a command from a set of commands, the set of commands comprising a write command and a plurality of non-write commands, and communicating the selected command to the memory, wherein a Hamming distance value calculated between a digital value indicating the write command and a digital value indicating any one of the multiple non-write commands is greater than 1.
In another embodiment, the invention provides a protocol operative within a memory system, comprising; defining unique digital values for each one of a set of commands, wherein the set of commands comprises a write command and a plurality of non-write commands, and wherein a Hamming distance value calculated between a digital value indicating the write command and a digital value indicating any one of the plurality of non-write commands is greater than 1.
Embodiments of the invention are described in relation to the accompanying drawings. Throughout the drawings and written description like reference numbers indicate like exemplary elements, components, and steps. In the drawings:
Embodiments of the invention will now be described with reference to the accompanying drawings. The invention may, however, be alternately and variously embodied and is not limited to only the illustrated embodiments. Rather, the illustrated embodiments are presented as teaching examples. The actual scope of the invention is defined by the claims that follow.
In one aspect, embodiments of the invention address the possible effect of data communication errors on memory system commands. Such commands may be communicated by discrete control signals, by data contained in a data packet, or other conventional means.
Many conventional memory systems sequentially execute commands only after finishing a corresponding EDC operation. When EDC data is communicated from a memory controller to a memory in relation to a command, an EDC operation must typically be performed to verify the accuracy of the data associated with the command (e.g., control data, address data, etc.) before the command can be executed in the memory. This sequential execution of the EDC operation followed by execution of the command operation ensures that the data received in relation to the command is accurate before the command is actually executed.
This straight-forward, sequential approach to command data verification followed by execution of the command ensures accuracy. It is also likely to create a command execution bottleneck in the memory. For example, if transient noise is apparent on a channel connecting the memory controller and memory, a group of commands sequentially received by the memory may include noise-induced data errors. The EDC operations necessarily used in a conventional memory system to identify and/or correct these errors may be relatively lengthy in their respective or cumulative execution. As a result, data throughput in the conventional memory system may be markedly reduced.
Against this backdrop, embodiments of the invention recognize practical execution distinctions between certain types of memory system operations. For example, embodiments of the invention recognize a clear distinction between “write operations” and “non-write operations.” For purposes of this description, a “write operation” is any operation that has the intended capability of altering one or more data value(s) in memory. This definition is broader than the phrase “write operation” used in the context of conventional RAM devices, for example. It also includes the program/write operations used in the context of conventional flash memory devices. Clearly, these types of conventionally understood operations—which are specifically ascribed to the alteration of data stored in memory—are subsumed within the working definition of the term “write operation.” However, other operations are also included within this definition. For example, erase operations performed in many kinds of nonvolatile memories are also intentionally capable of altering one or more data value(s) in memory, and as such, fall within the working definition of a write operation.
In contrast, other memory system operations, hereafter individually and collectively referred to as “non-write operations,” are not intentionally capable of altering one or more data value(s) in memory. Conventional read operations applied to both volatile and nonvolatile memories are examples of a non-write operation, since their execution is not intended to alter any data value stored in memory. So too, refresh operations applied to conventional DRAM devices are another example of a non-write operation, as are conventional chip select operations, etc.
In the description that follows, each operation, whether a write operation or a non-write operation, is assumed to be associated with one or more commands communicated from a memory controller to a memory. Some operations in certain memory systems may be executed in response to a single command indicated by a unitary block of command data. Other operations may be executed in response to multiple commands indicated over a sequence or collection of data blocks containing related command data. For purposes of clarity throughout this description, such distinctions in the execution details of various commands will not be further referenced, and each memory system operation will be said to be performed in response to “a command”, recognizing that said command may be indicated using many different data configurations and/or data transmission timing characteristics. Thus, in view of the foregoing, a write operation will be said to be executed in a memory in response to a “write command,” and a non-write operation will be said to be executed in a memory in response to a “non-write command.”
Commands and data associated with commands may be variously “communicated” within the context of embodiments of the invention between a memory controller and a memory. Such communication may be wireless and/or hardwired in its nature. Both standard and customized data communication protocols are contemplated by embodiments of the invention.
Commands and related memory system data (e.g., address data, write data, read data, etc.) may be communicated between a memory controller and a memory in a data packet or in some other non-packet form (i.e., using conventional serial or parallel data communication techniques). Data packets may be variously defined in their structure and communication timing. Embodiments of the invention contemplate both standard and customize data packet definitions. Thus, while certain embodiments of the invention have particular application and benefit to memory systems using packetized data, the subject invention is not limited to only such systems.
Commands, whether packetized or un-packetized, must be “decoded” within a memory before (or as part of) being executed using memory-based resources (e.g., hardware and/or software resources available to the memory). Packet decoding is a well understood concept. Since one common purpose for packetizing data is the multiplexing of corresponding signals lines, some partitioning and subsequent decoding of the data bits contained in a data packet are typically required. However, the term “decoding” as used in this description has a broader definition and subsumes all forms of digital data and/or analog signal manipulation, interpretation, and/or conversion.
In many instances, memory system data and/or signals may be identified as internal verses external for purposes of description. Terms like internal data, internal address, internal command, etc., indicate data operable or meaningful within a memory of the memory system. For example, in practical applications, memory controllers manufactured by a number of vendors may be used in memory systems in conjunction with memories manufactured by different vendors. It is common for various memory controller commands to be “decoded” into internal commands executable within a memory. Sometimes the format and definition of external and internal commands, or external and internal data may be identical. However, some decoding, conversion or interpretation of a command or data received from a memory controller may be performed by a memory. This is also true of certain control signals, such as clock signals. However, data/clock timing conversions, data type conversions, signal level conversions, command interpretations and decoding, etc. between the memory controller and memory are deemed to be conventional in their nature and are assumed to vary with memory system design.
Once a command and any related data has been received in memory, successfully identified by decoding, and verified by an applied EDC operation (if required), it may be executed using hardware and/or software resources in memory. The actual execution of commands in a particular memory system is a matter of design choice and is also deemed conventional for purposes of this description.
In the illustrated embodiments that follow, a generic memory core is indicated as a primary hardware resource for storing data. While DRAM examples have been selected to illustrate various embodiments of the invention, those of ordinary skill in the art will recognize that other memory types might be used, such as a SRAM, or a flash (NAND or NOR) memory. For the sake of clarity, conventional memory core operations will be assumed in order to avoid repetition of extraneous and conventionally understood detail related to the actual mechanisms whereby data storage (writing or programming) and data retrieval (reading) are accomplished. Thus, row and column addressing techniques, applied word line and bit line voltages, applied sequences of data operations, applied sequences of signal line and control signals, etc., will vary with the specific type and design of particular memory cores.
Embodiments of the invention also recognize that memory cores are commonly associated with a number of related circuits, such as decoders, voltage generators, redundant memory circuits, etc. These conventionally associated peripheral circuits are largely omitted from the illustrated embodiments that follow, since their design, use, and incorporation are deemed to be conventionally understood.
In recognition of the distinct operating implications between write operations and non-write operations, embodiments of the invention variously relate to a command protocol, as well as a memory within a memory system using the command protocol. In one embodiment, a command protocol designed in accordance with the dictates of the present invention further recognizes that erroneously communicated (and/or erroneously interpreted) write operations have a particularly high likelihood of causing major operating difficulties within a memory system. For example, where one or more data bits indicating a non-write command is changed during communication due to channel noise effects, the resulting collection of received data bits may actually be interpreted by the memory as a write command. Where such an erroneously indicated write operation is executed, critical data stored in memory may be overwritten and lost. In contrast, the erroneous interpretation and execution of a non-write command, while annoying in its delay upon overall memory system operations, is nonetheless unlikely to result in the corruption of stored data.
The pernicious effects of an errant write operation being executed in a memory system is one of the drivers motivating incorporation of EDC capabilities within memory systems. As noted above, EDC capabilities incorporated within a memory system may vary in their complexity, sophistication, and application. But regardless of the nature and sophistication of an EDC capability, it imposes material costs (operating speed, layout space, system complexity, etc.) on the design and operation of a memory system. Such costs tend to escalate with increasingly sophistication of the EDC capability. Thus, in so far as the simplicity and operating efficiency of an EDC capability can be increased, the easier and less costly its incorporation will be within a memory system. Ideally, a memory system would be sufficiently insulated from data communication errors such that no EDC capability would be required. However, in many memory system applications this ideal is unlikely to be met in the near future.
In view of the foregoing, the table of exemplary memory system commands shown in
However, in the example of
The possibility of this type of errantly communicated and/or erroneously interpreted command happening may be expressed mathematically by a calculated value known as a “Hamming distance value.” The last row of the table shown in
A Hamming distance value may be calculated in one approach by performing an exclusive-OR (XOR) operation on two digital values expressed on two strings of binary data (i.e., strings of binary command data indicating the respective memory system commands). As is well understood by those skilled in the art, an XOR operation may be implemented by simply adding the corresponding bits of the two data strings using a binary add operation without carry. This type of operation effectively identifies a number of corresponding bits between the two data strings that have different values. Consider the following example in which six (6) of the corresponding bit pairs between Data Strings “A” and “B” have different logic values. The calculated Hamming distance value (H) for this example is therefore six (6).
In similar fashion, the Hamming distance value between the active command (ACT) and the write command (WR) in the example of
Defining a memory system command protocol that increases the Hamming distance value between each write command and all non-write commands has the effect of establishing an interpretation/execution guard band of sorts around the write command(s). In other words, each increment in the Hamming distance value calculated between each write command and all non-write commands has the effect of increasing (by a squared factor, N2) the possibility of the non-write command being errantly communicated and/or erroneously interpreted as a write command.
Thus, using the example command set shown in
A memory system command protocol defining each write command as having a minimal Hamming distance value greater than 1 may be adapted for use in all kinds of memory systems. However, for purposes of illustration several exemplary memory systems will now be described. A comparative conventional memory system will first be described with reference to
The conventional memory system illustrated in
The external clock generated by memory controller 10 may be used to define uniform timing intervals during which data packets are communicated between memory controller 10 to memory 20.
Each of the comparative memory system embodiments presented herein might further incorporate EDC capabilities. If this were the case, the constituent command/address packets would also include EDC data. However, the illustrated memory system examples described herein are assumed to not include EDC capabilities for purposes of descriptive clarity. Those of ordinary skill in the art will recognize, however, that this need not be the case, and that memory systems (packetized and non-packetized) incorporating all manner of EDC capabilities may benefit from use of a memory system command protocol according to embodiments of the invention. Additional examples of memory systems incorporating EDC capabilities and likely benefiting from a memory command protocol according to the invention may be found in commonly assigned U.S. patent application [Attorney Docket Number SEC.2000] filed concurrently with the subject application, the subject matter of which is hereby incorporated by reference.
The exemplary embodiment of the invention illustrated in
In the illustrated example, memory controller 100 receives various requests from a host device CPU as well as related data. In response, memory controller 100 generates an external clock signal, a command/address (C/A) packet and, as required, write data. The command data and address data forming the C/A packet may be generated in a C/A encoding block (CAEP). Thereafter, a C/A transmission block (CATP) places the resulting C/A packet on the C/A bus of channel 130. The write data when needed is generated by a data processing block (DPP) and communicated via a write data bus (WD) to memory 200. Each of these constituent memory controller components may be implemented using conventional designs, and may be operated according to conventional techniques, except as otherwise required below.
At memory 200, the received C/A packet may be decoded by a C/A receiver 22 (C/A_R) into an internal command (com) and an internal address (add) in relation to the external clock signal or an internal clock signal derived from the external clock signal. The internal command (com) is applied to a command decoder 24 (CD) where it is decoded to produce one or more internal command signals applied to a memory core 32. The internal address (add) is applied to an address decoder 26 (AD) which in turn generates row and/or column address voltages applied to select specific memory cell(s) MC in memory core 32. The operation of memory core 32, the decoding of the internal command and internal address, as well as the generation and application of internal command signals, row and column address signals, word and bit line voltages, etc., may all be conventionally implemented.
Memory 200 also includes a data input buffer 28 receiving write data from the write data bus (WD) and thereafter communicating the write data to memory core 32 in response to a write enable signal (/WR) and a data output buffer 30 receiving read data from memory core 32 and thereafter communicating the write data to memory controller 100 in response to a read enable signal (/RE).
Within the exemplary memory system of
The same five (5) memory system commands (e.g., ACT, PRE, WR, RE and REF) indicated by the data packet of
The embodiment illustrated in
Once the one or more C/A/W packets are received in a C/A/W receiver 22′ (C/A/W_R) and decoded into one or more internal commands (com), an internal address, and internal write data (wd). The internal write data (wd) is applied to write data input buffer 28′ in response to a write enable signal /WR provided by command decoder 24.
With this particular channel configuration 131 including a single C/A/W bus communicating command, address, and write data, it may be necessary to communicate this block of data in multiple data packets. For example, if we assumed the same channel resources and time frame considerations assumed for the memory systems of
Each of the first and second data packets indicates its nature by also including a first transmitted data/command bit (DC). This bit identifies the data packet as either containing command/address data or containing write data. In response to the data/command bit, packet receiver 22′ in memory 210′ knows how to handle the constituent data. For example, in the embodiment of
Subsequently (or previously), when packet receiver 22′ receives the second data packet it generates internal write data (wd). The internal write data (wd) is communicated from packet receiver 24′ to a write data buffer 28′ (DB) where it is temporarily stored.
However, the embodiment illustrated in
The foregoing embodiments have been presented in a description that highlights certain functions and capabilities apparent in various memory controller and memory designs. The individual circuits identified in the foregoing embodiments in relation to these functions and capabilities should not be construed in an overly literal sense. Those of ordinary skill in the art will recognize that the functions and capabilities may be implemented in a great variety of different circuit combinations. So combinations and/or partitions of the exemplary circuits discussed above may be made. Alternately, software may be used to implement one or more of these functions and capabilities. Whether implemented in hardware and/or software, and whether implementing a packetized or non-packetized data communication arrangement, the memory may be generally said to include a decoding/execution block capable of receiving a memory system command, decoding the command, and executing the command in relation to one or more constituent memory cores.
Indeed, one or more controller(s), microprocessor(s), or Application Specific Integrated Circuits (ASICs) and related software may be used to implement the functions and capabilities ascribed to one or more of the circuits described above in the context of
Up to this point in the description, the exemplary embodiments have assumed a memory capable of being implemented in a single integrated circuit (IC), (e.g., a DRAM, SRAM, or flash memory). However, embodiments of the invention are by no means limited to single IC memory implementations, or implementations where one memory controller is matched with only one unitary memory device.
As described above, one important driver in the design and implementation of contemporary memory systems is size. The memory system of
Memory controller 10″ is connected to memory 20″ via channel 13″ which includes a number of high-speed point-to-point interconnections. Such high-speed point-to-point interconnections are highly susceptible to variances in the length of their constituent signal lines. Thus, extended lateral arrangements of memory elements are difficult to implement. However, vertically stacked arrangements of memory elements allow relatively short point-to-point interconnections between memory controller 10″ and memory 20″. Further vertical interconnections between stacked memories in each memory element may be configured according to conventionally understood techniques. For example, in one embodiment, a lowermost memory element may route (or re-drive) received packets to other memories in the stack in accordance with a received address. Each memory element may include its own error detector, or a commonly accessed (and potentially more powerful) error detector 28″ may be provided on memory 20″.
In another embodiment of the invention, a smart card may be implemented as a “memory” associated with a smart card terminal acting as a “memory controller” within the vernacular of the foregoing description. That is, commands communicated from the smart card terminal to the smart card may be handled in a manner consistent with the foregoing embodiments.
Indeed, many products incorporating contemporary and emerging memory systems will benefit from a memory system command protocol that provides a reduced possibility of non-write operations being errantly communicated and/or erroneously interpreted and executed as write operations. Indeed, this improved immunity to data error is not limited to only a distinction between write and non-write commands. Minimum Hamming distance values may be established in a set of memory system commands between any type of critical command and other commands deemed less critical in their execution.
When a memory system command protocol according to an embodiment of the invention is used, it may be possible to eliminate the need for a DEC capability within the memory system. Alternately, an existing EDC capability may be used with greater reliability. Alternately, a memory system might be provisioned with a less capable (smaller and more streamed-lined in operation) EDC capability while retaining an acceptable level of error immunity.
It should be noted that while the invention has been taught in the context of selected embodiments and detailed examples teaching the making and use of the invention, the invention is not limited to only the disclosed embodiments. Those of ordinary skill in the art will recognize that various modifications and alterations may be made to these embodiments without departing from the scope of the invention as defined by the following claims.
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US8045405 *||Jul 18, 2007||Oct 25, 2011||Samsung Electronics Co., Ltd.||Memory system, memory device and command protocol|
|Cooperative Classification||G06F11/1076, G06F11/1008|
|European Classification||G06F11/10M, G06F11/10R|