WO2015098707A1 - 情報検索機能を備えたメモリ、その利用方法、装置、情報処理方法。 - Google Patents
情報検索機能を備えたメモリ、その利用方法、装置、情報処理方法。 Download PDFInfo
- Publication number
- WO2015098707A1 WO2015098707A1 PCT/JP2014/083607 JP2014083607W WO2015098707A1 WO 2015098707 A1 WO2015098707 A1 WO 2015098707A1 JP 2014083607 W JP2014083607 W JP 2014083607W WO 2015098707 A1 WO2015098707 A1 WO 2015098707A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- memory
- information
- logical
- data
- word
- Prior art date
Links
- 230000015654 memory Effects 0.000 title claims abstract description 159
- 230000006870 function Effects 0.000 title claims description 63
- 230000010365 information processing Effects 0.000 title claims description 42
- 238000003672 processing method Methods 0.000 title claims description 7
- 238000000034 method Methods 0.000 title description 10
- 210000004027 cell Anatomy 0.000 claims description 14
- 239000004065 semiconductor Substances 0.000 claims description 6
- 210000000352 storage cell Anatomy 0.000 claims description 5
- 230000005055 memory storage Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 16
- 238000004364 calculation method Methods 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 10
- 238000001514 detection method Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 238000007418 data mining Methods 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 230000008707 rearrangement Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000012805 post-processing Methods 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- LFYJSSARVMHQJB-QIXNEVBVSA-N bakuchiol Chemical compound CC(C)=CCC[C@@](C)(C=C)\C=C\C1=CC=C(O)C=C1 LFYJSSARVMHQJB-QIXNEVBVSA-N 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
- G06F16/90339—Query processing by using parallel associative memories or content-addressable memories
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C15/00—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C15/00—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores
- G11C15/04—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores using semiconductor elements
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C7/00—Arrangements for writing information into, or reading information out from, a digital store
- G11C7/10—Input/output [I/O] data interface arrangements, e.g. I/O data control circuits, I/O data buffers
- G11C7/1006—Data managing, e.g. manipulating data before writing or reading out, data bus switches or control circuits therefor
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/28—Constructional details of speech recognition systems
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C15/00—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores
- G11C15/04—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores using semiconductor elements
- G11C15/043—Digital stores in which information comprising one or more characteristic parts is written into the store and in which information is read-out by searching for one or more of these characteristic parts, i.e. associative or content-addressed stores using semiconductor elements using capacitive charge storage elements
Definitions
- the present invention relates to a memory having an information search function, a method of using the same, an apparatus, and an information processing method.
- in-memory database technology is attracting attention as a technology for processing big data at high speed.
- One of them is data mining or the like that makes a large amount of data resident in a semiconductor memory (in-memory database) to speed up data access and to search for target information at high speed.
- the CPU Since the current computer is a computer that handles all information processing, the CPU is forced to undertake both the processing that the CPU is good at and the processing that the CPU is not good at.
- the data on the memory is like a playing card turned upside down, and there is no other way than searching for information by turning (accessing) each piece (one address, one address).
- the CPU performs information processing that sequentially searches for information on the memory and finds specific information, the amount of information processing becomes extremely large and the waiting time increases. This is the fate of a Neumann computer and the bus bottleneck.
- CAM associative memory
- the content addressable memory (CAM) requires a completely parallel comparison circuit, which increases the circuit size, and the parallel circuit constituting the comparison circuit consumes a large current. Therefore, at present, the content addressable memory (CAM) is only used in a special field such as a communication router in which a super-high-speed search is indispensable.
- the present invention can realize a memory based on a new concept that enables retrieval of big data at a speed comparable to an associative memory (CAM) by incorporating a very small amount of circuit into a general memory. It is the purpose.
- CAM associative memory
- a memory having an information narrowing detection function is a memory that is good at logical product operations such as pattern matching.
- PCT / JP2013 / 059260 a memory with set operation function, can expand and expand the concept of memory with information narrowing detection function above, and freely perform logical product operation, logical sum operation, logical negation operation, etc. It is memory that can.
- the disclosures of the two applications are hereby incorporated in their entirety by this reference.
- the memory 101 of the present invention can be applied to the above two prior inventions.
- the memory with an arithmetic function is intended to improve chip efficiency by providing an arithmetic circuit for each block as shown in the figure.
- the size of the arithmetic circuit is smaller than the case where all the memories have individual arithmetic circuits, the arithmetic efficiency is deteriorated, and the efficiency of the chip is low, resulting in a cost problem.
- information retrieval which is the biggest problem of information processing by the CPU, that is, information processing is complicated, it cannot be handled by anyone other than an expert, the burden on the CPU is large, the peripheral circuit is complicated, and the power consumption is large.
- CAM associative memory
- the memory according to claim 1 is capable of reading and writing information.
- This memory has a memory cell structure of 1 word bit width of n and N word addresses, that is, N * n bits.
- a set of n-bit logical operation units (3)
- the memory cell information of one word bit width n repeatedly selected and specified from among the N word addresses is inputted (assigned) in parallel to the one set of n-bit logical operation units.
- a function for performing logical operation (4)
- the logical operation unit according to claim 2 wherein the logical operation unit is a logical storage, a logical product, a logical sum, a logical negation, or an exclusive logic for each bit of the memory storage cell information of which the one word bit width is n.
- the data value complete match and the data value range search are performed by any combination operation of logical storage, logical product, logical sum, logical negation, and exclusive logic.
- a memory having an information retrieval function is assumed.
- Claim 6 is a memory having an information search function according to claim 1, which is integrated with another type of semiconductor device such as a CPU.
- a memory having an information search function according to the first aspect of the present invention which is mounted on an FPGA.
- the memory having the information search function is used in (9) a serial connection, a parallel connection, or a serial parallel connection, and (2) a hierarchical connection (1) or (2) connection.
- a method of using a memory having the information retrieval function described in 1 is used.
- a tenth aspect of the present invention is an apparatus including the memory according to the first aspect.
- An information processing method wherein a predetermined operation result is obtained by repeating logical operation of information of 1 bit of memory storage cell information in a memory, logical sum, logical negation, exclusive logic, and a combination thereof.
- FIG. 1 is a configuration diagram of a general memory.
- FIG. 2 is a configuration diagram of a memory having an information search function.
- FIG. 3 is an example of a document search using a memory having an information search function (Example 1).
- FIG. 4 is an example of search for completely matched data by a memory having an information search function (Example 2).
- FIG. 5 is an example of range data search by a memory having an information search function (Example 3).
- FIG. 6 shows an example of series-parallel connection of memories having an information search function (Example 4).
- FIG. 7 is an example of hierarchical connection of memories having an information search function (Example 5).
- FIG. 1 is a configuration diagram illustrating a general memory according to an embodiment.
- the memory 100 in FIG. 1 does not include functional circuits such as an address decoder and a data bus, and has a configuration in which information data can be freely written to and read from this memory.
- One word has a width 103 of n bits and N word addresses.
- the memory cell 102 is composed of N ⁇ n bit cells having 104, and generally word addresses 1 to N can be selected and designated from the outside by means such as an address decoder.
- Information processing by the current CPU is that the data width 103 of the memory 100 is a constant data width such as 8 bits, 16 bits, 32 bits, and the address space of the given memory such as 1M address or 1G address is used when searching for information data.
- the CPU sequentially accesses the address, reads data, and performs sequential processing.
- the information processing by the memory of the present invention is based on the idea of reversing the common sense of the data width and address concept of the above general memory structure and database table structure, and is based on parallel logical operations in 1-bit units. It is.
- FIG. 2 shows an example of the configuration of a memory having the information search function of this embodiment.
- One word has a width 103 of n bits and N words
- a memory cell 102 having an address 104 and made up of N ⁇ n bit cells is provided, and a word address from 1 to N can be selected and designated 110 from the outside.
- the width 103 of one word n bit corresponds to the number of records (n) in the database, and one record is arranged in a column, and N of the word address 104 is easily understood as a structure corresponding to a field of one record.
- this memory is a data table of 1 record with N bits and n records.
- the logical operation unit 105 provided in parallel with the n-bit storage cell 102, which is selected and specified 110 by the word address 104 in the row direction (horizontal direction in this figure) of the memory, stores the word address 104 which is selected and specified 110.
- a circuit capable of logical storage 116 for each cell 102 bits, logical product 112, logical sum 113, logical negation (NOT) 114, exclusive logic 116, and a combination thereof can be freely specified.
- a calculation result output function 106 such as a priority address encoder output circuit is provided to output the calculation result of the logic calculator 105.
- Most of the memory is the memory cell itself, and only a small part is the function of the logical operation unit 105 and the operation result output 106. Therefore, these functions can be used for a database by incorporating these functions in a very small area of a general memory. It can be a large-capacity memory.
- a plurality of sets of logical operation units 105 may be prepared by making it possible to select and specify a plurality of word addresses 110 simultaneously.
- the memory capacity per semiconductor die (chip) at the present time is about 8 Gbit.
- the main memory 101 having a word width of 8 Kbit when the word address is 1M, and the word width of 1 Mbit when the word address is 8K.
- the memory 101 having any combination of vertical and horizontal is realized, such as the main memory 101. This memory 101 is effective for searching for all kinds of information, especially for searching for big data, data mining, genome analysis, and the like, and examples of using this memory will be described.
- An example of our daily information search is the Internet search, and the concept is realized by narrowing down information by keywords. For example, by providing keywords such as “information processing”, “information search”, and “CPU”, narrowing down is performed and an appropriate Internet site is found.
- FIG. 3 shows an example of document retrieval using a memory having an information retrieval function.
- the word addresses from 1 to N are words such as “information processing”, “information retrieval”, “patent”, “CPU”, etc., and one vertical column of word width n is one record.
- word addresses 1 to N constituting one record correspond to fields. That is, if there is at least one character such as “information processing”, “information retrieval”, “patent”, “CPU” in one document, “1” is written in the corresponding memory cell (field) ( “0” is omitted, and so on). Therefore, in this example, N vocabularies and n books (n records) are registered as a database.
- Document records containing the vocabulary of either (information processing) at word address 18 or “information retrieval” at word address 5 (logical sum (OR)) are 3, 4, 5, 13, 14, 16, 19, 21 and 25.
- the word address 24 is “patent”, and the documents that do not include the “patent” vocabulary are 4, 8.11, 16, 22, 25, and the above calculation results
- document records 3, 4, 5, 13, 16, 16, 19, 21, 25 and the logical product (AND) result of performing the logical product operation of the logical negation operation 114 are 4, 16, 25.
- the final winning remaining document 107 is the record 16 It has become.
- the document 16 is (a document including a vocabulary of either “information processing” or “information retrieval”) ⁇ (a document not including a patent vocabulary) ⁇ (a document including a CPU vocabulary).
- the above results may be read sequentially from the operation result output 106 such as a priority address encoder output circuit.
- the CPU can detect the target information from this memory without searching all the information of the entire memory space only by performing the word address selection designation 110 and the calculation designation 111 in the memory 101.
- FIG. 4 shows an example of searching for exact match data using a memory having an information search function.
- the word address 10 As the most significant bit (MSB) and the word address 17 as the least significant bit (LSB) are allocated to the field. Since it is 8-bit data, it is possible to store 256 kinds of data. By appropriately selecting eight word addresses from word address 10 to word address 17, it is possible to discriminate between 256 kinds of data based on complete coincidence. Any search is possible. For example, when the data value “10” and the binary number “000001010” are searched for a complete match, the word address 10 is calculated as the most significant bit (MSB) and the word address 17 is calculated eight times to the least significant bit (LSB). What is necessary is just to detect some data.
- MSB most significant bit
- LSB least significant bit
- the calculation is performed in order from the MSB word address 10, where “0000” digit “0” digit is logical negation and “1” digit case Is positive logic, and two records 13 and 25 that have repeatedly won 8 logical product operations (winning operations) have a data value “10”. If an addition function or a subtraction function is incorporated in the logical operation unit 105 and performed in parallel, four arithmetic operations on record data can be performed.
- FIG. 5 shows an example of range data search using a memory having an information search function.
- the complete match of the data value “10” is obtained.
- searching for the data value “10” or more as shown in FIG.
- the logical value of the lower 4 bits of the word address 15 and 16 and the logical operation of the word address 14 are used to obtain the data value “10” or more and less than “16”, and the record with the previous data value “16” or more is obtained.
- a record having a data value of “10” or more can be searched.
- a record having a data value of “10” or more is negated, a record of less than “10”, that is, “9” or less is detected.
- the same 1-bit operation may be repeated.
- the above calculation is the result of processing all records in parallel in about 10 times. Realizing a range search from perfect match by just double the above if the data value is 16 bits and 4 times if the data value is 32 bits. I can do it. Further, even when the data width is increased from 8 bits to 9 bits or 10 bits, it is very simple, and it is not always necessary that word addresses are continuous, and it is possible to realize the data width to 17 bits or 33 bits without a sense of incongruity.
- this memory can be freely assigned in the field from 1 bit data with or without to the range search of arbitrary data width.
- the information is narrowed down by searching for “male from Chiba prefecture, working in Tokyo, height 170cm to 175cm, 50 to 60 years old”, etc.
- the major feature is that only necessary records can be detected.
- the calculation result output 106 such as a priority address encoder output circuit can be divided into several blocks so that it can be read out in units of blocks.
- Data can be written and read using a fixed-width data bus, and can be configured to be written and read serially in a first-in first-out (FIFO) interface in the vertical and horizontal directions. You can also
- Fig. 6 shows an example when this memory is connected in series and parallel.
- the system can be expanded both in the vertical direction (word address direction) and in the horizontal direction (data width direction), so that the system can be expanded very simply and the system can be made persistent.
- the required capacity in the horizontal direction (data width direction) is also determined in the vertical direction (word address direction).
- the number of word addresses is several hundred thousand. However, in the case of personal information, it is sufficient to have several K to several K words per person.
- the access speed varies depending on the memory element, for example, if the speed of one logical operation is 10 ns, if the time for searching for information is several hundred ns to microseconds and 1 s, 100,000 operations can be realized.
- This technology is capable of finding the target information in a fixed time of several hundreds of nanoseconds to microseconds or 1 millisecond, regardless of any big data. It is a feature.
- Fig. 7 shows an example of hierarchical connection of memories having an information search function.
- the memory 101 having the information search function shown at the top of the figure is used as a master, corresponding to each record, and provided with a sub-information search function storing more detailed data for each record.
- the memory 101 is configured to be searchable. In particular, in the case of big data, it is possible to deal with any size database by using such a hierarchical database.
- this memory 101 In order to construct a database using this memory 101, it is possible to use only how records and fields are allocated, and how the logical operation unit 105 is designated for operation 111. Therefore, this memory 101 is used. By using it, the database concept itself such as a conventional search algorithm such as SQL is made unnecessary. Information search using a CPU has various usage techniques for reducing the burden on the CPU, and a binary search is a typical example.
- This algorithm is a standard technology for information processing as a technology that can extremely reduce the number of searches for information data.
- data values are written in a data table on a memory
- prior arrangement is made such that data is arranged in order from small data to large data. It is necessary to rearrange the data on the memory (data maintenance) whenever preparation is necessary and data increases or decreases.
- this algorithm reduces the burden on the CPU when searching for a specific data value, but the burden on pre-processing and data maintenance before that is quite small.
- Patent No. 4588114 A memory having an information narrowing-down detection function is a memory that is good at AND operations such as pattern matching.
- PCT / JP2013 / 059260 memory with set operation function expands and expands the concept of memory with information narrowing detection function, and can perform logical product operation, logical sum operation, logical negation operation, etc. It is memory that can.
- a memory having an information narrowing detection function and a memory having a set operation function can be realized.
- this method can realize a large equivalent associative memory (CAM) although the speed of the parallel operation itself is somewhat reduced, so that the processing time for the entire information retrieval is greatly reduced. I can do it.
- Such a memory 101 is extremely effective for genome analysis.
- Human DNA and the like is big data having several G of base information, and analysis of such big data requires an extremely long time. If base information of several G is stored in the memory 101 at once, DNA analysis becomes faster and more accurate than a supercomputer.
- the current information processing is such that the CPU sequentially accesses the address, reads the data, and sequentially performs the information processing with a constant data width such as the data width of the memory 100 such as 32 bits, 64 bits, and 128 bits.
- the wider the data width (bus width) the higher the efficiency of information processing, but there are limits to the expansion of the data bus width, such as the increase in the number of input / output pins of the device and the increased wiring burden on the printed circuit board on which the device is mounted. There is.
- the feature of the memory 101 of the present invention is that the conventional memory structure and database information processing are reversed in the horizontal and vertical directions, and parallel operations are performed for each 1 bit in the memory. 102 If information can be processed in parallel, the number of information processing operations will be extremely small and efficient. Furthermore, the characteristics of the memory device, that is, a large number of memory cells in parallel inside the memory chip without providing information to the outside It is utilized that it is possible to perform parallel operation by inputting (substituting) into a parallel logic unit.
- the algorithm of the memory 101 can be easily mounted on the FPGA.
- This memory is an in-memory database and has a self-search function that does not rely on the CPU.
- a device configuration that integrates this memory and the CPU is also effective.
- the associative memory has been known as a device with a high-speed computing function from the past, and various applications including various knowledge processing have been studied, but it cannot have a large storage capacity, and has low power consumption. It is used only for some purposes such as communication routers because of its large size.
- the present invention is a new type of information processing memory that has a large capacity and an information retrieval speed comparable to that of an associative memory (CAM), it can be widely used for data mining, data analysis, and knowledge processing of various databases and big data. Is possible.
- CAM associative memory
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Memory System (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
情報の読み出し書き込みが可能なメモリであって
(1)このメモリは1ワード bit幅がnでNワードアドレス、つまりN*n bitの記憶セル構造とし
(2)このメモリ全体に1組n bitの論理演算器
(3)前記Nワードアドレスの中から繰り返し選択指定される前記1ワード bit幅がnの記憶セル情報を前記1組n bitの前記論理演算器に並列に入力(代入)し論理演算する機能
(4)前記論理演算器の内容を出力する機能
以上を具備することを特徴とする情報検索機能を備えたメモリとする。
前記論理演算器は前記入力(代入)される前記1ワードbit幅がnのメモリ記憶セル情報同士のbit毎の、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算が出来る構成である事を特徴とする請求項1記載の情報検索機能を備えたメモリとする。
前記論理演算器に、シフトレジスタ機能を具備することを特徴とする請求項1記載の情報検索機能を備えたメモリとする。
前記、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算により、データ値の完全一致並びにデータ値の範囲検索を行うことを特徴とする請求項1記載の情報検索機能を備えたメモリとする。
前記、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算により、データ値の加減演算を行うことを特徴とする請求項1記載の情報検索機能を備えたメモリとする。
CPUなど他の種類の半導体デバイスと一体化されたことを特徴とする請求項1記載の情報検索機能を備えたメモリとする。
FPGAに実装された事を特徴とする請求項1記載の情報検索機能を備えたメモリとする。
データベースのレコードを前記1ワードbit幅nのいずれか1列に割り付けし、1レコードのフィールド総bit数を前記ワードアドレス数のNとするデータベースとすることを特徴とする請求項1記載の情報検索機能を備えたメモリの使用方法とする。
前記情報検索機能を備えたメモリを
(1)直列、並列、もしくは直並列に接続
(2)階層的接続
(1)または(2)の接続で使用することを特徴とする請求項1記載の情報検索機能を備えたメモリの使用方法とする。
請求項1記載のメモリを含んだ装置とする。
メモリ内部でメモリ記憶セル情報の1bit同士の情報の論理積、論理和、論理否定、排他論理、その組み合わせの論理演算を繰り返し所定の演算結果を得ることを特徴とする情報処理方法とする。
前記1bit同士の情報を並列に前記論理演算することを特徴とする請求項11記載の情報処理方法とする。
またPCT/JP2013/059260号 集合演算機能を備えたメモリは以上の情報絞り込み検出機能を備えたメモリの概念を拡大発展させて、論理積演算、論理和演算、論理否定演算などを自由に行うことができるメモリである。
101 情報検索機能を備えたメモリ
102 記憶セル
103 ワード幅
104 ワードアドレス
105 論理演算器
106 演算結果出力
107 演算結果
110 (ワードアドレス)選択指定
111 演算指定
112 論理積
113 論理和
114 論理否定
115 排他論理
116 論理記憶
Claims (12)
- 情報の読み出し書き込みが可能なメモリであって
(1)このメモリは1ワード bit幅がnでNワードアドレス、つまりN*n bitの記憶セル構造とし
(2)このメモリ全体に1組n bitの論理演算器
(3)前記Nワードアドレスの中から繰り返し選択指定される前記1ワード bit幅がnの記憶セル情報を前記1組n bitの前記論理演算器に並列に入力(代入)し論理演算する機能
(4)前記論理演算器の内容を出力する機能
以上を具備することを特徴とする情報検索機能を備えたメモリ。 - 前記論理演算器は前記入力(代入)される前記1ワードbit幅がnのメモリ記憶セル情報同士のbit毎の、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算が出来る構成である事を特徴とする請求項1記載の情報検索機能を備えたメモリ。
- 前記論理演算器に、シフトレジスタ機能を具備することを特徴とする請求項1記載の情報検索機能を備えたメモリ。
- 前記、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算により、データ値の完全一致並びにデータ値の範囲検索を行うことを特徴とする請求項1記載の情報検索機能を備えたメモリ。
- 前記、論理記憶、論理積、論理和、論理否定、排他論理その何れの組合せ演算により、データ値の加減演算を行うことを特徴とする請求項1記載の情報検索機能を備えたメモリ。
- CPUなど他の種類の半導体デバイスと一体化されたことを特徴とする請求項1記載の情報検索機能を備えたメモリ。
- FPGAに実装された事を特徴とする請求項1記載の情報検索機能を備えたメモリ。
- データベースのレコードを前記1ワードbit幅nのいずれか1列に割り付けし、1レコードのフィールド総bit数を前記ワードアドレス数のNとするデータベースとすることを特徴とする請求項1記載の情報検索機能を備えたメモリの使用方法。
- 前記情報検索機能を備えたメモリを
(1)直列、並列、もしくは直並列に接続
(2)階層的接続
(1)または(2)の接続で使用することを特徴とする請求項1記載の情報検索機能を備えたメモリの使用方法。 - 請求項1記載のメモリを含んだ装置。
- メモリ内部でメモリ記憶セル情報の1bit同士の情報の論理積、論理和、論理否定、排他論理、その組み合わせの論理演算を繰り返し所定の演算結果を得ることを特徴とする情報処理方法。
- 前記1bit同士の情報を並列に前記論理演算することを特徴とする請求項11記載の情報処理方法。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP14874258.8A EP3091448A4 (en) | 2013-12-23 | 2014-12-18 | Memory equipped with information retrieval function, method for using same, device, and information processing method |
JP2015554805A JP5981666B2 (ja) | 2013-12-23 | 2014-12-18 | 情報検索機能を備えたメモリ、その利用方法、装置、情報処理方法。 |
CN201480070692.5A CN105900085B (zh) | 2013-12-23 | 2014-12-18 | 具备信息检索功能的存储器及其利用方法、装置、信息处理方法 |
US15/107,565 US9627065B2 (en) | 2013-12-23 | 2014-12-18 | Memory equipped with information retrieval function, method for using same, device, and information processing method |
US15/425,555 US20170147712A1 (en) | 2013-12-23 | 2017-02-06 | Memory equipped with information retrieval function, method for using same, device, and information processing method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2013264763 | 2013-12-23 | ||
JP2013-264763 | 2013-12-23 |
Related Child Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/107,565 A-371-Of-International US9627065B2 (en) | 2013-12-23 | 2014-12-18 | Memory equipped with information retrieval function, method for using same, device, and information processing method |
US15/425,555 Continuation US20170147712A1 (en) | 2013-12-23 | 2017-02-06 | Memory equipped with information retrieval function, method for using same, device, and information processing method |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015098707A1 true WO2015098707A1 (ja) | 2015-07-02 |
Family
ID=53478566
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2014/083607 WO2015098707A1 (ja) | 2013-12-23 | 2014-12-18 | 情報検索機能を備えたメモリ、その利用方法、装置、情報処理方法。 |
Country Status (5)
Country | Link |
---|---|
US (2) | US9627065B2 (ja) |
EP (1) | EP3091448A4 (ja) |
JP (2) | JP5981666B2 (ja) |
CN (1) | CN105900085B (ja) |
WO (1) | WO2015098707A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017014322A1 (ja) * | 2015-07-23 | 2017-01-26 | 井上 克己 | 情報検出プロセッサ及びその方法 |
WO2018097317A1 (ja) * | 2016-11-28 | 2018-05-31 | 井上 克己 | データ比較演算プロセッサ及びそれを用いた演算方法 |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
RU2652501C1 (ru) * | 2017-06-14 | 2018-04-26 | Виктория Владимировна Олевская | Модуль поиска блока информации по входным данным |
KR102631380B1 (ko) * | 2018-05-17 | 2024-02-01 | 에스케이하이닉스 주식회사 | 데이터 연산을 수행할 수 있는 다양한 메모리 장치를 포함하는 반도체 시스템 |
JP6986309B1 (ja) * | 2021-09-22 | 2021-12-22 | 明男 三水 | データ処理装置、データ処理方法、及びプログラム |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH01297724A (ja) * | 1988-05-25 | 1989-11-30 | Nec Corp | 学習型文字列検索装置と同装置の制御方式 |
JP4588114B1 (ja) | 2010-02-18 | 2010-11-24 | 克己 井上 | 情報絞り込み検出機能を備えたメモリ、その使用方法、このメモリを含む装置。 |
WO2011102432A1 (ja) * | 2010-02-18 | 2011-08-25 | Inoue Katsumi | 情報絞り込み検出機能を備えたメモリ、このメモリを用いた情報検出方法、このメモリを含む装置、情報の検出方法、メモリの使用方法、およびメモリアドレス比較回路 |
WO2013147022A1 (ja) * | 2012-03-28 | 2013-10-03 | Inoue Katsumi | 集合演算機能を備えたメモリ及びこれを用いた集合演算処理方法 |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5325500A (en) * | 1990-12-14 | 1994-06-28 | Xerox Corporation | Parallel processing units on a substrate, each including a column of memory |
JP3595565B2 (ja) * | 1993-12-24 | 2004-12-02 | シャープ株式会社 | 半導体メモリセル回路およびメモリセルアレイ |
JP2000067573A (ja) | 1998-08-19 | 2000-03-03 | Mitsubishi Electric Corp | 演算機能付きメモリ |
EP1182577A1 (en) * | 2000-08-18 | 2002-02-27 | SER Systeme AG Produkte und Anwendungen der Datenverarbeitung | Associative memory |
JP2006127460A (ja) * | 2004-06-09 | 2006-05-18 | Renesas Technology Corp | 半導体装置、半導体信号処理装置、およびクロスバースイッチ |
US8341362B2 (en) * | 2008-04-02 | 2012-12-25 | Zikbit Ltd. | System, method and apparatus for memory with embedded associative section for computations |
CN101577757A (zh) * | 2009-03-05 | 2009-11-11 | 上海闻泰电子科技有限公司 | 实现短信检索功能的方法 |
US20120117317A1 (en) * | 2009-08-20 | 2012-05-10 | Rambus Inc. | Atomic memory device |
-
2014
- 2014-12-18 CN CN201480070692.5A patent/CN105900085B/zh active Active
- 2014-12-18 JP JP2015554805A patent/JP5981666B2/ja active Active
- 2014-12-18 WO PCT/JP2014/083607 patent/WO2015098707A1/ja active Application Filing
- 2014-12-18 EP EP14874258.8A patent/EP3091448A4/en active Pending
- 2014-12-18 US US15/107,565 patent/US9627065B2/en active Active
-
2016
- 2016-07-28 JP JP2016148348A patent/JP6229024B2/ja active Active
-
2017
- 2017-02-06 US US15/425,555 patent/US20170147712A1/en not_active Abandoned
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH01297724A (ja) * | 1988-05-25 | 1989-11-30 | Nec Corp | 学習型文字列検索装置と同装置の制御方式 |
JP4588114B1 (ja) | 2010-02-18 | 2010-11-24 | 克己 井上 | 情報絞り込み検出機能を備えたメモリ、その使用方法、このメモリを含む装置。 |
WO2011102432A1 (ja) * | 2010-02-18 | 2011-08-25 | Inoue Katsumi | 情報絞り込み検出機能を備えたメモリ、このメモリを用いた情報検出方法、このメモリを含む装置、情報の検出方法、メモリの使用方法、およびメモリアドレス比較回路 |
WO2013147022A1 (ja) * | 2012-03-28 | 2013-10-03 | Inoue Katsumi | 集合演算機能を備えたメモリ及びこれを用いた集合演算処理方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP3091448A4 |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017014322A1 (ja) * | 2015-07-23 | 2017-01-26 | 井上 克己 | 情報検出プロセッサ及びその方法 |
WO2018097317A1 (ja) * | 2016-11-28 | 2018-05-31 | 井上 克己 | データ比較演算プロセッサ及びそれを用いた演算方法 |
JP6393852B1 (ja) * | 2016-11-28 | 2018-09-19 | 井上 克己 | データ比較演算プロセッサ及びそれを用いた演算方法 |
Also Published As
Publication number | Publication date |
---|---|
JP5981666B2 (ja) | 2016-08-31 |
JP6229024B2 (ja) | 2017-11-08 |
US20160322105A1 (en) | 2016-11-03 |
CN105900085A (zh) | 2016-08-24 |
EP3091448A4 (en) | 2017-12-06 |
JP2017016668A (ja) | 2017-01-19 |
US9627065B2 (en) | 2017-04-18 |
JPWO2015098707A1 (ja) | 2017-03-23 |
EP3091448A1 (en) | 2016-11-09 |
US20170147712A1 (en) | 2017-05-25 |
CN105900085B (zh) | 2019-08-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6229024B2 (ja) | 情報検索機能を備えたメモリ、その利用方法、装置、情報処理方法。 | |
US20160140243A1 (en) | Scoped search engine | |
Kaplan et al. | PRINS: Processing-in-storage acceleration of machine learning | |
EP3907627A1 (en) | Technologies for performing stochastic similarity searches in an online clustering space | |
US8943091B2 (en) | System, method, and computer program product for performing a string search | |
Lee et al. | Application codesign of near-data processing for similarity search | |
Yavits et al. | GIRAF: General purpose in-storage resistive associative framework | |
US20200265045A1 (en) | Technologies for refining stochastic similarity search candidates | |
Chowdhury et al. | A DNA read alignment accelerator based on computational RAM | |
CN113627618A (zh) | 使用启用列读取的存储器来实行随机稀疏提升和普罗克拉斯提斯正交稀疏散列的技术 | |
CN108268596B (zh) | 搜索存储在存储器中的数据的方法和系统 | |
Garzón et al. | AIDA: Associative in-memory deep learning accelerator | |
WO2016199808A1 (ja) | メモリ型プロセッサ、メモリ型プロセッサを含んだ装置、その使用方法。 | |
WO2015094721A2 (en) | Apparatuses and methods for writing masked data to a buffer | |
US20200410039A1 (en) | Data comparison arithmetic processor and method of computation using same | |
US20230385258A1 (en) | Dynamic random access memory-based content-addressable memory (dram-cam) architecture for exact pattern matching | |
Gagie et al. | Compressed spaced suffix arrays | |
Yavits et al. | PRINS: resistive CAM processing in storage | |
Zeng | FPGA-based high throughput merge sorter | |
JPWO2014168199A1 (ja) | 論理演算方法および情報処理装置 | |
JP5208117B2 (ja) | 表形式データを操作するマルチコア対応データ処理方法、マルチコア型処理装置、及び、プログラム | |
JP2018169636A (ja) | データ検索装置、データ検索lsi。 | |
Yavits | Will computing in memory become a new dawn of associative processors? | |
JP2015090567A (ja) | インデキシング方法、インデキシング装置、集積回路 | |
Stefan et al. | The Connex Engine: An in-memory device for fast string operations |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14874258 Country of ref document: EP Kind code of ref document: A1 |
|
DPE1 | Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101) | ||
ENP | Entry into the national phase |
Ref document number: 2015554805 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15107565 Country of ref document: US |
|
REEP | Request for entry into the european phase |
Ref document number: 2014874258 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2014874258 Country of ref document: EP |