WO2019005888A1 - SYSTEMS AND METHODS FOR TRANSFORMING LARGE DATA TO A REDUCED SIZE REPRESENTATION - Google Patents
SYSTEMS AND METHODS FOR TRANSFORMING LARGE DATA TO A REDUCED SIZE REPRESENTATION Download PDFInfo
- Publication number
- WO2019005888A1 WO2019005888A1 PCT/US2018/039620 US2018039620W WO2019005888A1 WO 2019005888 A1 WO2019005888 A1 WO 2019005888A1 US 2018039620 W US2018039620 W US 2018039620W WO 2019005888 A1 WO2019005888 A1 WO 2019005888A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- bit
- bit pattern
- data
- test
- pattern
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
- G06F12/10—Address translation
- G06F12/1027—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB]
- G06F12/1045—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB] associated with a data cache
- G06F12/1054—Address translation using associative or pseudo-associative address translation means, e.g. translation look-aside buffer [TLB] associated with a data cache the data cache being concurrently physically addressed
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0751—Error or fault detection not based on redundancy
- G06F11/0754—Error or fault detection not based on redundancy by exceeding limits
- G06F11/076—Error or fault detection not based on redundancy by exceeding limits by exceeding a count or rate limit, e.g. word- or bit count limit
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/58—Random or pseudo-random number generators
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
- H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
- H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
- H03M7/55—Compression Theory, e.g. compression of random number, repeated compression
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2212/00—Indexing scheme relating to accessing, addressing or allocation within memory systems or architectures
- G06F2212/40—Specific encoding of data in memory or cache
- G06F2212/401—Compressed data
Definitions
- Embodiments of the present invention related generally to data compression and decompression techniques.
- the present invention provides a system comprising at least one processor with one or more cores; and memory including, in part, instructions to configure the at least one processor to perform a method, the method comprising: receiving a data set of a data length; determining a bit pattern of the data set; generating a reference set of bit patterns, the set being of a set length, the set length being equivalent to the data length, the set of bit patterns including every possible different bit pattern from all zeros to all ones; determining a first test bit pattern using a first bit pattern generation function applied to test data; determining a distance between the first test bit pattern and the bit pattern of the data set using a location of the first test bit pattem and a location of the bit pattern of the data set, the locations being relative to the reference set of bit patterns; iterating the first test pattern generation function in a direction of the bit pattem of the data set and combining the first test pattern generation function with at least one second test pattern generation function to determine a second test bit pattern applied to the test data;
- FIG. 1 is a block diagram depicting a system for transforming data between its larger representation and its smaller representation according to some embodiments.
- FIG. 2 is a block diagram that illustrates a computer system upon which one or more of data transformation operations described herein may be implemented according to some embodiments.
- FIGs 3(a), 3(b) and 3(c) show multiple possible implementations of a Data Compiler (DC) and Turing Dedekind device (TD).
- DC Data Compiler
- TD Turing Dedekind device
- FIG. 4 shows a vector providing both a visual representation of the combinatorial complexity of any size N bit pattern for all N which also provides a template illustrating N different but related sets of operations performed by both the DC at load time and the TD at run time.
- FIG. 5 illustrates a HCSSFHIL showing that it is composed of two or more LOOPS each of which represents generating functions that produced output bit patterns containing the characteristics which qualify it as a HCSSFHIL as represented by HCSSFHIL's descriptive name. These loops are connected by one or more different operations that can represent addition or any other function.
- FIG. 6 illustrates how sequentially iterating HCSSFHIL's are produced by the DC with each HCSSFHIL outputting a particular J value which then gets altered to produce the next HCSSFHIL which outputs the next J value.
- FIG. 7 illustrates how the DC utilizes metadata items including but not limited to virtual memory addresses and or time stamps to transform some or all of its parameters possibly including representation of its error term to lower the spatial footprint and or increase the calculating power of the parameters that the TD will actually store and utilize via a method in which one or more of these parameters are themselves recalculated from bit patterns contained in virtually memory addresses, time stamps and or other metadata and or from functions of bit patterns contained in virtually memory addresses, time stamps and or other metadata.
- metadata items including but not limited to virtual memory addresses and or time stamps to transform some or all of its parameters possibly including representation of its error term to lower the spatial footprint and or increase the calculating power of the parameters that the TD will actually store and utilize via a method in which one or more of these parameters are themselves recalculated from bit patterns contained in virtually memory addresses, time stamps and or other metadata and or from functions of bit patterns contained in virtually memory addresses, time stamps and or other metadata.
- FIG. 8 illustrates how the TD implements the analogous operations at run time.
- FIG. 9 represents the iterative steps described in these specifications showing step by step how the DC starting form an input bit pattern k executes a series of iterative steps the produce the DCs output in the form of the kernel that it hands to the TD.
- This invention provides a game changing solution to a series of long standing, seemingly intractable problems that have served to drastically limit the potential positive performance impact of forty plus generations of exponential hardware improvements under Moore's Law dating back to the invention of the Integrated Circuit and the appearance of the first microprocessors. These problems include the problem of big data together with the problem of network congestion and our inability to find useful applications for multi core architecture.
- Each bit pattern's informational content is first transformed via being spatially reduced to a highly compressed kernel, via the mechanism contained in a Data Compiler (hereinafter DC) as described below, in order to one or more times be subsequently recreated in its original format via recalculation from the contents of that kernel, via the mechanism contained in a Turing Dedekind device (hereinafter TD) as is likewise described below.
- DC Data Compiler
- TD Turing Dedekind device
- Embodiments of the current invention operate upon and allow for the recalculation of, while obviating the need to store, bit patterns.
- Each such bit partem has a specific length, measurable in bits, so that any and all specific bit patterns k, of any and all specific lengths n, as measured by the number of bits it contains represents the fundamental unit of data, regardless of the content, purpose and/or utility of that data, that the methods and apparatuses of the current invention operate upon - thereby first enabling the DC to create the necessary kernel, including parameters and code, which occupies a far smaller spatial footprint, that the TD is then able to execute in order to rapidly losslessly in a practical fail safe manner recalculate the specific n bit long bit partem k that was originally input into the DC, thereby enabling the aforementioned desirable results.
- bit patterns For the sake of simplicity, we shall for the purposes of this exposition consider such bit patterns to consist of classical mutually exclusive on or off binary digits traditionally represented by zeros and ones, but the methods and apparatuses of this invention are equally applicable of being implemented on data stored in other, not necessarily yet fully practical, but potentially vastly more efficient basic data representational units including but not limited to q bits and e bits with the ability to know how to implement this invention in multiple possible alternate data representational realms, including but not limited to the realm of quantum computing utilizing (non-binary, non-simple On is not off) q bits and e bits, that is capable of being accomplished by those trained in such state of the art practice.
- the methods and apparatuses of this invention are capable of being implemented in hardware, software or via other mechanisms such as firmware and/or microcode and/or any combination thereof.
- multiple embodiments of this invention in various possible combinations in part or all of some or any of the following media / devices / implementations, etc., in hardware software, firmware, and/or microcode, etc., can also accommodate varying possible degrees from minimal to extensive partial human intervention and possible human or other non-computerized contributions, so that any possible partial human and/or other noncomputerized contribution and/or other input, or intervention (or none at all) into any device, apparatus and/or method implemented in any possible practical workable combination of hardware, software, firmware, microcode, and/or any other computerized implementations methods and/or apparatuses of at least some part or all of the body and contents of the embodiments of the present invention as described herein (including or not including any possible or potential partial human
- bit patterns sent over communications lines can be further distinguished from each other by such characteristics as frequency, signal intensity, etc., in addition to the 2N possible distinct informational content states based solely on bit patterns, and embodiments of the present invention likewise provides means of distinguishing the informational content contained within two or more separate but bit-content-identical bit patterns via the use of such factors as the virtual memory addresses at which these bit patterns are stored and the time stamps indicating the time certain operations were performed on these bit patterns.
- each such configuration can contain 2N distinct bit patterns so that the actual cumulative total of distinct separate bit patterns capable of being produced by all these possible configurations taken jointly will be determined by such factors as the number of duplicate results and whether specific operations are commutative, under conditions where there is clearly no reason to expect that the cumulative sum total number of all such possible results won't exceed 2N distinct bit patterns.
- the ability to create n separate say trillion-to-one compression ratios by taking successive powers of a very small number of bits or via achieving similar ratios via executing other loops with large indexes and a small number of parameters with a short spatial footprint creates a situation a) in which all numeric differences between, and/or other simple functional combinations between any two or more such highly compressed bit strings produce a far larger new set of almost equally highly compressed bit strings, b) in which this or other similar operations as described in a) above can be performed recursively leading to a further combinatorial explosion of highly compressed bit patterns with each iteration, and c) in which the additional operation of any possible function that can be described in, for example, one kilobyte creates 2**8000 separate times the sum total of bit patterns produced by the operations described in a) and b) above, each of which will manifest a roughly one-billion-to-one compression ratio.
- FIG. 1 is a block diagram depicting an example of a data transformation system 100 according to some embodiments.
- the data transformation scheme 100 includes a large data store 102, data transformation devices 104, a smaller data representation 106, and
- the large data store 102 represents data in non-compressed, unaltered form.
- the large data store may include raw data.
- the large data store 102 is to be processed by a processor of a computer system for carrying out intended operations based on the large data store 102.
- the large data store 102 may include user data, such as documents, images, pictures, video, and audio, etc., to be processed by the processor and presented to users.
- the large data store 102 may include program files including instructions to be executed by the processor to perform an operation instructed by the instructions.
- the large data store 102 may include system data for managing hardware and software resources managed by the computing system.
- the large data store 102 may be stored in memory, persistent storage, or in combinations thereof.
- the large data store 102 may be consumed in portions by the computing system.
- the data transformation devices 104 includes a device (e.g., a DC) for data transformation to the smaller representation and a device (e.g., a TD) for re-transformation back to the original large data store.
- a device e.g., a DC
- a device e.g., a TD
- transformation techniques described in this paper are employed to generate the smaller data representation 106.
- the smaller data representation 106 is re-transformed back into the large data store 102 using the reverse of those techniques.
- a largely expandable data transformation technique corresponding to the transformation technique is carried out using specific data transformation algorithms. The specific techniques for data transformation and re- transformation are described herein.
- the smaller data representation 106 represents data in small footprint that can be reproduced back to the large data store.
- the smaller data representation may include transformed data as a result of data transformation through the data transformation devices 104 and a smaller data representation obtained from data sources 108.
- the destination 108 may represent a datastore (e.g., memory, persistent storage, removable storage, or cache) in which the smaller data representation 106 generated as a result of data transformation is received.
- the data source 108 may represent a datastore (e.g., memory, persistent storage or cache) from which smaller data representation 106 to be re-transformed is obtained.
- the destinations/sources 108 may include a local storage drive, including internal storage devices integrated with a computing system and removable external storage devices connected to the computing system. The internal storage devices may include optical drive, hard disk, and solid state drive, and so on.
- the removable external storage devices may include a memory card (e.g., SD card), a memory stick (e.g., USB memory), and so on.
- the destinations/sources 108 may include external computing systems connected to a local computing system that performs the data transformation over a computer network.
- the external computing systems may include cloud server systems, server systems dedicated for specific entity, client computing systems, and so on.
- the cloud server systems may include cloud server systems, server systems dedicated for specific entity, client computing systems, and so on.
- destination/source 108 may include a data transport layer capable of transmitting or receiving the smaller data representation.
- FIG. 2 is a block diagram that illustrates a computer system 200 upon which one or more of data transformation operations according to the embodiments described herein may be implemented.
- the computer system 200 includes a bus 202 or other communication mechanism for communicating data, one or more hardware processors 204 coupled with the bus 202 for processing data.
- the hardware processor(s) 204 may be, for example, one or more general purpose microprocessors or one or more specific purpose microprocessors. As noted herein, the processors 204 may be a multi-core processor.
- the hardware processor(s) 204 may be configured to execute a data transformation engine 218 therein, which will be described below.
- the computer system 200 also includes a main memory 206, such as a random access memory (RAM), cache and/or other dynamic storage devices, coupled to the bus 202 for storing temporary data and instructions to be executed by the processor(s) 204.
- the instructions include the data transformation engine 218 configured to cause the processor(s) 204 to transform data (original) into the smaller data representation and re-transform the smaller data representation into the large data store according to data transformation algorithms described in the present disclosure.
- the main memory 206 also may be used for temporarily storing re-transformed data generated by executing the data transformation engine 218, such that the processor(s) 204 can process the re-transformed data.
- Such instructions when stored in storage media accessible to the processor(s) 204, render the computer system 200 into a special-purpose machine that is customized to perform the operations specified in the instructions.
- the word “engine” may constitute either software engines (e.g., code embodied on a machine-readable medium) or hardware engines.
- a “hardware engine” is a tangible unit capable of performing certain operations and may be configured or arranged in a certain physical manner.
- one or more computer systems e.g., a standalone computer system, a client computer system, or a server computer system
- one or more hardware engines of a computer system e.g., a processor or a group of processors
- software e.g., an application or application portion
- the computer system 200 further includes a read only memory (ROM) 208 or other static storage device coupled to the bus 202 for storing static information and instructions for the processor(s) 204.
- ROM read only memory
- the data transformation engine 218 may be loaded to the main memory 206, when the processor(s) 204 carries out a data transformation operation.
- a storage device 210 such as a magnetic disk, optical disk, or USB thumb drive (Flash drive), etc., is provided and coupled to the bus 202 for storing data and instructions.
- the storage device 210 includes a large data store 220.
- the computer system 200 may be coupled via the bus 202 to output device(s) 212, such as a cathode ray tube (CRT) or LCD display (or touch screen), for displaying information to a computer user.
- output device(s) 212 such as a cathode ray tube (CRT) or LCD display (or touch screen)
- Input device(s) 214 are coupled to the bus 202 for communicating information and command selections to the processor(s) 204.
- a cursor controller 216 such as a mouse, a trackball, or cursor direction keys for communicating direction information and command selections to the processor(s) 204 and for controlling cursor movement on the output device(s) 212.
- This input device 214 typically has two degrees of freedom in two axes, a first axis (e.g., x) and a second axis (e.g., y), that allows the device to specify positions in a plane.
- a first axis e.g., x
- a second axis e.g., y
- the same direction information and command selections as cursor control may be implemented via receiving touches on a touch screen without a cursor.
- the computing system 200 may include a user interface module to implement a GUI that may be stored in a mass storage device as executable software codes that are executed by the computing device(s).
- This and other modules may include, by way of example, components, such as software components, object-oriented software components, class components and task components, processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables.
- the computer system 200 may implement the techniques described herein using customized hard-wired logic, one or more ASICs or FPGAs, firmware and/or program logic which in combination with the computer system causes or programs the computer system 200 to be a special-purpose machine. According to some embodiments, the techniques herein are performed by the computer system 200 in response to the processor(s) 204 executing one or more sequences of one or more instructions contained in main memory 206. Such instructions may be read into the main memory 206 from another storage medium, such as the storage device 210. Execution of the sequences of instructions contained in the main memory 206 causes the processor(s) 204 to perform the process steps described herein. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions.
- the computer system 200 also includes a communication interface 218 coupled to the bus 202.
- the communication interface 218 provides a two-way data communication coupling to one or more network links that are connected to one or more local networks.
- the communication interface 218 may be an integrated services digital network (ISDN) card, cable modem, satellite modem, or a modem to provide a data communication connection to a corresponding type of telephone line.
- ISDN integrated services digital network
- the communication interface 218 may be a local area network (LAN) card to provide a data communication connection to a compatible LAN (or WAN component to communicated with a WAN).
- LAN local area network
- Wireless links may also be implemented.
- the communication interface 218 sends and receives electrical, electromagnetic or optical signals that carry digital data streams representing various types of information.
- a network link typically provides data communication through one or more networks to other data devices.
- a network link may provide a connection through local network to a host computer or to data equipment operated by an Internet Service Provider (ISP).
- ISP Internet Service Provider
- the ISP in turn provides data communication services through the world wide packet data communication network now commonly referred to as the "Internet”.
- Internet Internet
- Local network and Internet both use electrical, electromagnetic or optical signals that carry digital data streams.
- the signals through the various networks and the signals on network link and through communication interface 218, which carry the digital data to and from computer system 200, are example forms of transmission media.
- the computer system 200 can send messages and receive data, in particular, smaller data representation, through the network(s), network link and the communication interface 218.
- a server might transmit a requested code for an application program through the Internet, the ISP, the local network and the communication interface 218.
- the data transformation devices 104 operates to transform and re-transform data as generally as described herein.
- the term "data compiler” is edifying in so far as optimizing compilers transform computer code from a more readily human understandable but far more highly machine inefficient format into a functionally equivalent far more highly computer efficient but roughly correspondingly far less easily humanly understandable format
- the Data Compiler (DC) as described herein perform an analogous functionality by transforming data in the form of highly computer inefficient bit patterns (that are extremely costly to store and move around) into a far smaller kernel that can fit entirely or largely into a machine's on chip cache that provide CPU, graphics, or other processing entity the ability to rapidly losslessly recalculate the original bit pattern that was originally input into the data compiler, while a corresponding Turing Dedekind device (TD) performs the requisite high speed lossless recalculation of this original bit pattern from the kernel that the DC produced.
- TD Turing Dedekind device
- the term “Turing Dedekind device” derives from the method the TD uses to perform this recalculation in a manner that can usefully be conceptualized by considering key methodological design features that it possesses, which are highly pronounced of both a Turing Machine and a Dedekind Cut.
- bit patterns such as j and k are recalculated are reminiscent of a Turing Machine in so far as the process by which the TD recalculates j and/or k is via an accumulation process that may or may not be monotonic - which corresponds to a Turing Machine's ability to move a tape either unidirectionally or back and forth - proceeding at a known rate - so that key to the ability of the TD to losslessly reproduce k most efficiently lies in the ability of the DC to derive increasingly more accurate approximations with smaller error terms, the most efficient of which will most typically be executed by the TD at run time which based on the accumulation rate required to reproduce j precisely repeatedly being known to the system, the method by which the TD stops accumulating when it has produced the bit pattern j corresponds conceptually to a Turing machine moving a tape unidirectionally or back and forth in the exact proper predetermined manner combined with knowing the right time to halt.
- the DC In contrast with the TD, which is the conceptually simpler of these two devices in so far as it "simply” executes the code against the kernel, both of which are handed to it by the DC, in order to recreate either some distinctly specified subset or the totality of the bit partem that was originally input into the DC, the DC needs to "perform the heavy lifting" including determining both the kernel and the code operating on it for each bit partem k it receives, which kernel and code the DC needs to design and create for the express purpose of eliminating the need to store bit pattern k which is now replaced with the ability of a TD to recalculate k, utilizing high speed processors, from a far smaller kernel which the DC needs to provide to the TD, together with the code, that DC likewise needs to provide which likewise exhibits a sufficiently small spatial foot print, that enables the TD to convert the highly compressed kernel which the DC has handed it back into the original bit pattern k in a vastly more machine efficient manner than the traditional legacy process of saving, storing and transporting and retriev
- kernel can be reasonably used to reference either the parameters and the code operating on it or alternatively solely the parameters that the code operates on without this potentially ambiguous dual meaning ever becoming a potential or actual source of confusion as long as it is made clear in each and every such specific instance, either explicitly, or implicitly via context, in which of these two distinct ways the term "kernel" is being used.
- microprocessor's on chip cache are microprocessor's on chip cache.
- Such efficacious loops include but are not limited to raising X to the N power recursively k times for potentially large k, periodic alterations of a loops inner contents via increasing decreasing or otherwise functionally transforming one or more parameters within the loop every k iterations including but not limited to, for example, the exact analogue to the generating function of the Mandelbrot Set performed in the real rather than the complex plane, any calculation modulo m, any average, weighted average, numeric difference between or other simple function of any combination of two or more generating functions, the inclusion of additional terms containing constants, in addition to various possible bit shifting operations, bit pattern inverting, cutting and/or splicing and/or concatenations of different bit pattern segments, in addition to performing other simple logical and/or other mathematical operations together with numerous other variants recognizable and/or discoverable to those trained in the art, all of which possibilities can be further expanded by the inclusion of even significantly large error terms
- HCSSFHIL pronounced HICKS-FILL or HICKSVILLE or LONG ISLAND
- LOOP a distinct calculation unit within a HCSSFHIL
- HCSSFHIL by definition can contain any number of LOOPS greater than one, the number of LOOPS in each HCSSFHIL in the preferred embodiment of this invention is determined by practical optimization efficiency considerations under circumstances in which more LOOPS per HCSSFHIL provides a more efficient targeting mechanism at the cost of more processing steps in the TD.
- any alternative function other than addition that combines multiple bit pattems produced by two or more LOOPS together to produce a typically significantly longer distinct bit partem whose numeric value corresponds to the numeric value of the bit pattern produced by the HCSSFHIL as a result of executing these one or more of these alternative, not necessarily additive functions against the contents of the output of the particular multiplicity of LOOPS inside each HCSSFHIL is capable of providing the connecting mechanism that the DC utilizes to turn multiple LOOPS into a HCSSFHIL.
- each bit pattern that supplies input to the DC that will subsequently be recalculated to provide the output produced by its corresponding TD can, but need not, be divided into two or more HCSSFHILs.
- the simplest way to divide a bit pattern into two or more HCSSFHILs to be followed by enabling the TD to reconnect multiple distinct smaller partial recalculated bit patterns at run time in order to losslessly recreate the original bit partem that was first input into the DC at load time from these separate parts is to first divide the entire original bit pattern that was first input into the DC into K separate N bit long bit patterns, each of which are then processed separately as established by the DC at load time in order to be recombined via concatenation at run time by the TD.
- any function that is utilized to first divide a bit partem into to two or more HCSSFHILs whose inverse operation is then employed to recombine the resultant output at run time in order to losslessly reproduce the original bit pattern may be used in an embodiment of the current invention.
- each loop internal to a specific HCSSFHIL in addition to the HCSSFHIL that contains each of these loops manifests its own specific ACCUMULATION RATE, known to the DC and expressible in terms of known first and second and possibly higher derivatives, based upon which the CD can always alter one or more parameters within its loops, including but not limited to the index that determines when a loop halts in addition to which the DC can always add new loops with specifically chosen parameters and/or it can remove one or more loops that that HCSSFHIL already contains in manner that can always be designed to insure that the newly altered bit pattern ja+1 that each successive newly designed "next iteration" HCSSFHIL outputs will contain an associated numeric value that is closer than the associated numeric value of its immediately prior "predecessor" bit pattern j a was to its target as determined by the unique associated value of the bit pattern k that that HCSSFHIL is attempting to best approximate as determined ultimately by both the unique associated input value corresponding
- bit pattern k containing a more random bit pattern includes both utilizing the increased "smoothness" of possible distribution outcomes produced by greatly reducing the likelihood that any two arbitrary bit patterns will proceed through the iterative processes described above at greatly different rates due to statistical anomalies associated with non-random bit patterns that may or may not impact positively or negatively the rate at which the aforementioned iterative process will be successfully concluded by the DC.
- the second potential advantage of initially randomizing the bit pattern k originally being input to the DC lies in taking positive advantage of statistical anomalies that will apply to an extremely small subset of the bit patterns which possess the requisite equivalence relationship with a given highly randomized input bit pattern k, the potential advantage of this technique being most clearly obvious in the extreme case where the original bit partem k occupies a sufficiently small spatial footprint and/or it can be sufficiently advantageously subdivided into such smaller segments and utilizing perhaps large expensive servers with massive
- the DC employs standard methods known to those trained in the art (e.g., when utilizing the experimental mathematical methods such as the integer relationship detection methods referenced previously) in order to uncover the function needed to effectively encode the Error Term as described above and likewise other such methods known to those trained in the art may be utilized by the DC in enabling embodiments of this invention in order to represent the Error Term directly, or indirectly via the use of a function, in a manner that makes use of other known methods to further reduce the size of the spatial footprint required to represent the Error Term inside the kernel.
- standard methods known to those trained in the art e.g., when utilizing the experimental mathematical methods such as the integer relationship detection methods referenced previously
- other such methods known to those trained in the art may be utilized by the DC in enabling embodiments of this invention in order to represent the Error Term directly, or indirectly via the use of a function, in a manner that makes use of other known methods to further reduce the size of the spatial footprint required to represent the Error Term inside the kernel.
- the DC may employ one or more iterations in which two or more successive ja's may temporarily move further away from k in order to find a path which under these specific circumstances known to the DC approaches k more quickly than it would if it had utilized a more direct route.
- Such methods include but are not at all limited to methods specified above that make use of the large quantity of precisely accurate statistical data that the DC has available to it, including for example both knowledge of each subsection of the bit partem originally input to it and different ways of slicing and dicing it capable of producing different associated numeric values and other such obviously useful logical bit-wise determination properties of each of its distinct separate segments under these multiple possible slicing and dicing operations together with complete knowledge of the numeric value of the distance between each of these parts and a large collection of known highly compressible small spatial footprint known generating functions whose varying accumulation rates are also known to the DC and which can be asymptotically goal directedly modified however the DC finds most useful as described above. But this is not nearly exhaustive of the types of tools that the DC has available at its disposal.
- Such means of representing informational content can frequently be represented via multiple pointers, mappings, and/or other content labeling, content referencing and/or content identifying denotations and/or procedures that serve substantially similar and/or identical useful functional purposes but whose spatial footprint is often quite different.
- a small series of time stamps and/or virtual memory addresses that are themselves 64 bit entities but which can be uniquely identified by their ordinal order or their position in a list and/or by some other pointer mechanism including for example by a recursive systems of pointers and/or mappings that can be used to uniquely identify specific data items where these pointing or other such functional entities can often be represented by different length pointers as a function for example of the number of different items being pointed to at different levels within a larger more complex hierarchy.
- This aspect of the present invention provides an ability to implement real time updates (a capability not present in traditional data compression) in so far as the final output value of a bit pattern is via dependent in part on such factors as virtual memory addresses and/or time stamps and/or functions thereof, in a deterministic manner known to the DC, making quick on the fly changes to certain segments of a large bit stream can be achieved by changing the values of the associated virtually memory address and/or time stamps and/or other metadata repositories and/or functions thereof in a manner that causes the TD to output the newly changed bit pattern.
- Such changes in for example 64 bit entities representing virtual memory addresses, time stamps or other metadata can by utilized for indexing purposes providing an ability to access part of the overall bit pattern starting or stopping at arbitrarily designated points. Note that this can be accomplished utilizing either a subset of the bit patterns in for example a 64 bit entity such as a virtual memory address or time stamp or some other metadata repository and/or it can be represented positionally and/or some combination thereof which can be useful for example in performing such indexing in a manner that accommodates two or more independent indexing criteria. Consider for example that it would require 14 bits to differentiate every second (which could be used to denote any arbitrarily chosen starting and/or stopping point) in a four and a half hour long movie.
- a hybrid mechanism could be utilized accommodating the ability to search positionally for a specific sub-location at some specific position within a cache line.
- positional information can be used for other purposes including more efficiently encoding information.
- end result of that computation will be order invariant permitting the order the DC places the units in according to some ordering principle (e.g., the associated numeric value of that subsection of the appropriate length within that binary bit pattern) and then encoding further information based upon which of the n! possible permutational orderings was in fact chosen.
- each bit pattern input into the DC is represented by one or more HCSSFHILs and each HCSSFHIL is represented by two or more LOOPS a case in which the entire input bit partem is capable of being recalculated from a single loop is also to be considered an embodiment of this invention.
- FIGs 3(a), 3(b) and 3(c) show multiple possible implementations of a DC and TD.
- FIG. 3(a) shows a DC and the corresponding TD residing mainly or totally in the on chip cache on a single computer in which the DC established (at load time) the processing operations and the data parameters that form the kernel that the TD will operate upon to rapidly losslessly reproduce the bit pattern originally input into the DC.
- FIG. 3(b) shows the same functionality achieved with the DC residing on one machine (e.g., a server) and while TD containing the kernel resides mainly or totally in the on chip cache of a separate machine (e.g., a client).
- FIG. 3(c) is distinguished from FIG. 3(b) in that the bit pattern that provides the input to the DC originates from a machine other than the server from another machine that may or may not be a client.
- FIG. 4 shows a vector providing both a visual representation of the combinatorial complexity of any size N bit pattern for all N which also provides a template illustrating N different but related sets of operations performed by both the DC at load time and the TD at run time.
- the vector itself contains a conceptual representation of every possible bit content that can be represented a size N bit pattern, e.g., it contains 2**8N elements starting with all zeros and ending with all ones in which the associated numeric value of each element is equal to the associated numeric value of the bit pattern immediately preceding it plus one.
- bit pattern represented some place on a line segment corresponding to this vector designated as k represents the particular bit partem being input to the DC at load time which likewise corresponds exactly to the bit pattern that is output by the TD at run time.
- Both the DC and the TD typically produce a series of other distinct size N bit patterns that are designed to increasingly approximate the numeric value of bit pattern k, each of which is designated by a particular Ja.
- the successive J values produced by the DC are typically altered iteratively whereas the last set of Ja's produced by the DC correspond to the Ja's that are utilized by the TD.
- HCSSFHIL shows that it is composed of two or more LOOPS each of which represents generating functions that produced output bit patterns containing the characteristics which qualify it as a HCSSFHIL as represented by HCSSFHIL's descriptive name. These loops are connected by one or more different operations that can represent addition or any other function.
- FIG. 6 illustrates how sequentially iterating HCSSFHIL's are produced by the DC with each HCSSFHIL outputting a particular J value which then gets altered to produce the next HCSSFHIL which outputs the next J value.
- FIG. 7 illustrates how the DC utilizes metadata items including but not limited to virtual memory addresses and or time stamps to transform some or all of its parameters possibly including representation of its error term to lower the spatial footprint and or increase the calculating power of the parameters that the TD will actually store and utilize via a method in which one or more of these parameters are themselves recalculated from bit patterns contained in virtually memory addresses, time stamps and or other metadata and or from functions of bit patterns contained in virtually memory addresses, time stamps and or other metadata.
- metadata items including but not limited to virtual memory addresses and or time stamps to transform some or all of its parameters possibly including representation of its error term to lower the spatial footprint and or increase the calculating power of the parameters that the TD will actually store and utilize via a method in which one or more of these parameters are themselves recalculated from bit patterns contained in virtually memory addresses, time stamps and or other metadata and or from functions of bit patterns contained in virtually memory addresses, time stamps and or other metadata.
- FIG. 8 illustrates how the TD implements the analogous operations at run time. As such, it represents the inverse operations illustrated in FIG. 7 so that the output of FIG. 8 corresponds exactly to the input of FIG. 7 and the input of diagram 6 corresponds exactly to the output of FIG. 7.
- FIG. 9 represents the iterative steps described in these specifications showing step by step how the DC starting form an input bit pattern k executes a series of iterative steps the produce the DCs output in the form of the kernel that it hands to the TD.
- Steps one and two representing initially randomizing bit pattern k through the application of for example traditional compression methods and procedures that the DC may or may not employ to optimize the first iteration of calculating its initial j value.
- the DC then iterates through the various steps illustrated here to provide a series of additional j values whose aim is to more closely approximate the original bit pattern k.
- the DC tests at each stage whether it is sufficiently close according to its internal testing criteria after which it may or may not implement the further steps illustrated in FIG. 7 to further improve the efficiency of the operations and or data that will form the kernel that the TD will implement including by still further lowering its spatial footprint.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
- Devices For Executing Special Programs (AREA)
- Complex Calculations (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Priority Applications (6)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020217026017A KR20210107136A (ko) | 2017-06-26 | 2018-06-26 | 큰 데이터를 더 작은 표현으로 변환하고 더 작은 표현을 원래의 큰 데이터로 재변환하기 위한 시스템들 및 방법들 |
| JP2019572780A JP6903773B2 (ja) | 2017-06-26 | 2018-06-26 | 大きなデータをより小さな表現に変換する、およびより小さな表現を当初の大きなデータに戻して再変換するためのシステムおよび方法 |
| KR1020207002460A KR102366069B1 (ko) | 2017-06-26 | 2018-06-26 | 큰 데이터를 더 작은 표현으로 변환하고 더 작은 표현을 원래의 큰 데이터로 재변환하기 위한 시스템들 및 방법들 |
| EP18824155.8A EP3646208A4 (en) | 2017-06-26 | 2018-06-26 | SYSTEMS AND PROCEDURES FOR TRANSFORMING LARGE DATA INTO A SMALLER REPRESENTATION |
| CN201880051817.8A CN111357203A (zh) | 2017-06-26 | 2018-06-26 | 用于将大数据变换成较小的表示的系统和方法 |
| KR1020247003615A KR102782120B1 (ko) | 2017-06-26 | 2018-06-26 | 큰 데이터를 더 작은 표현으로 변환하고 더 작은 표현을 원래의 큰 데이터로 재변환하기 위한 시스템들 및 방법들 |
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US201762525068P | 2017-06-26 | 2017-06-26 | |
| US62/525,068 | 2017-06-26 | ||
| US201862658432P | 2018-04-16 | 2018-04-16 | |
| US62/658,432 | 2018-04-16 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2019005888A1 true WO2019005888A1 (en) | 2019-01-03 |
Family
ID=64742164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/US2018/039620 Ceased WO2019005888A1 (en) | 2017-06-26 | 2018-06-26 | SYSTEMS AND METHODS FOR TRANSFORMING LARGE DATA TO A REDUCED SIZE REPRESENTATION |
Country Status (6)
| Country | Link |
|---|---|
| US (3) | US10621108B2 (cg-RX-API-DMAC7.html) |
| EP (1) | EP3646208A4 (cg-RX-API-DMAC7.html) |
| JP (2) | JP6903773B2 (cg-RX-API-DMAC7.html) |
| KR (3) | KR102366069B1 (cg-RX-API-DMAC7.html) |
| CN (1) | CN111357203A (cg-RX-API-DMAC7.html) |
| WO (1) | WO2019005888A1 (cg-RX-API-DMAC7.html) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112578188B (zh) * | 2020-11-04 | 2023-07-07 | 深圳供电局有限公司 | 电气量波形的生成方法、装置、计算机设备和存储介质 |
| IL318736A (en) | 2021-06-16 | 2025-03-01 | Conti Innovation Center Llc | Mechanically stacked solar transmissive cells or modules |
| WO2023027087A1 (ja) | 2021-08-25 | 2023-03-02 | ユニチカ株式会社 | 軟磁性ナノワイヤーおよびそれを含む塗料ならびにそれを塗布してなる積層体 |
| KR20250127636A (ko) | 2024-02-19 | 2025-08-26 | 삼성전자주식회사 | 메모리 장치 및 메모리 장치를 포함하는 컴퓨팅 시스템 |
| US12414402B1 (en) | 2025-01-03 | 2025-09-09 | Conti Innovation Center, Llc | Optimizing cadmium (CD) alloy solar cells with sputtered copper-dopped zinc telluride (ZNTE:CU) back contacts in the presence of hydrogen |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5838832A (en) * | 1995-04-03 | 1998-11-17 | Iterated Systems, Inc. | Method and system for representing a data set with a data transforming function and data mask |
| US20020154770A1 (en) * | 1999-11-09 | 2002-10-24 | Short Kevin M. | Method and apparatus for chaotic opportunistic lossless compression of data |
| US20150169462A1 (en) * | 2012-07-05 | 2015-06-18 | Blancco Oy Ltd | Apparatus, a system, a method and a computer program for erasing data stored on a storage device |
Family Cites Families (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5359551A (en) * | 1989-06-14 | 1994-10-25 | Log Point Technologies, Inc. | High speed logarithmic function generating apparatus |
| US5544338A (en) * | 1992-12-31 | 1996-08-06 | International Business Machines Corporation | Apparatus and method for raster generation from sparse area array output |
| US5751740A (en) * | 1995-12-14 | 1998-05-12 | Gorca Memory Systems | Error detection and correction system for use with address translation memory controller |
| JP2974280B2 (ja) * | 1996-09-11 | 1999-11-10 | 日本電気通信システム株式会社 | ネットワーク接続のブリッジ装置における仮想グループ情報管理方法 |
| US6415408B1 (en) * | 1999-11-03 | 2002-07-02 | Unisys Corporation | Multi-stage algorithmic pattern generator for testing IC chips |
| US20030105945A1 (en) * | 2001-11-01 | 2003-06-05 | Bops, Inc. | Methods and apparatus for a bit rake instruction |
| ATE370495T1 (de) * | 2000-06-20 | 2007-09-15 | Univ New Hampshire | Komprimieren und dekomprimieren von audiodateien unter verwendung eines chaotischen systems |
| US20080214300A1 (en) * | 2000-12-07 | 2008-09-04 | Igt | Methods for electronic data security and program authentication |
| US7116663B2 (en) * | 2001-07-20 | 2006-10-03 | Pmc-Sierra Ltd. | Multi-field classification using enhanced masked matching |
| WO2004032435A1 (ja) * | 2002-10-03 | 2004-04-15 | In4S Inc. | ビットストリングの照合方法および装置 |
| TW591388B (en) * | 2003-02-21 | 2004-06-11 | Via Tech Inc | Memory address decoding method and related apparatus by bit-pattern matching |
| JP2005051479A (ja) | 2003-07-28 | 2005-02-24 | Dainippon Printing Co Ltd | 時限暗号化/復号システム |
| US7266661B2 (en) * | 2004-05-27 | 2007-09-04 | Silverbrook Research Pty Ltd | Method of storing bit-pattern in plural devices |
| US7558941B2 (en) * | 2005-06-30 | 2009-07-07 | Intel Corporation | Automatic detection of micro-tile enabled memory |
| US7805588B2 (en) | 2005-10-20 | 2010-09-28 | Qualcomm Incorporated | Caching memory attribute indicators with cached memory data field |
| US7389453B2 (en) * | 2005-10-20 | 2008-06-17 | Jon Udell | Queuing methods for distributing programs for producing test data |
| US8225297B2 (en) * | 2005-12-07 | 2012-07-17 | Microsoft Corporation | Cache metadata identifiers for isolation and sharing |
| JP2007282024A (ja) | 2006-04-10 | 2007-10-25 | Hitachi Kokusai Electric Inc | 無線通信装置 |
| US8541724B2 (en) * | 2006-09-29 | 2013-09-24 | Lone Star Ip Holdings, Lp | Small smart weapon and weapon system employing the same |
| US8127200B2 (en) | 2006-12-24 | 2012-02-28 | Sandisk Il Ltd. | Flash memory device and system with randomizing for suppressing errors |
| US8140908B2 (en) * | 2007-06-22 | 2012-03-20 | Microsoft Corporation | System and method of client side analysis for identifying failing RAM after a user mode or kernel mode exception |
| US7707371B1 (en) * | 2007-09-10 | 2010-04-27 | Cisco Technology, Inc. | Storage area network (SAN) switch multi-pass erase of data on target devices |
| US8860594B2 (en) * | 2012-05-17 | 2014-10-14 | Brilliant Points, Inc. | System and method for digital signaling |
| US9159423B1 (en) * | 2012-09-28 | 2015-10-13 | Cadence Design Systems, Inc. | Robust erase page detection logic for NAND flash memory devices |
| JP6197876B2 (ja) | 2013-10-03 | 2017-09-20 | 富士通株式会社 | プログラム、符号化装置、及び符号化方法 |
| US9244861B2 (en) * | 2013-11-13 | 2016-01-26 | Violin Memory Inc. | Efficient caching system |
| US9563251B2 (en) * | 2013-12-28 | 2017-02-07 | Intel Corporation | Representing a cache line bit pattern via meta signaling |
| US9690278B1 (en) * | 2014-04-10 | 2017-06-27 | Altera Corporation | Method and apparatus for high-level programs with general control flow |
| US10110558B2 (en) | 2014-04-14 | 2018-10-23 | Cavium, Inc. | Processing of finite automata based on memory hierarchy |
| US9300320B2 (en) * | 2014-06-27 | 2016-03-29 | Qualcomm Incorporated | System and method for dictionary-based cache-line level code compression for on-chip memories using gradual bit removal |
| EP3304746B1 (en) | 2015-05-21 | 2021-04-07 | Zeropoint Technologies AB | Methods, devices and systems for hybrid data compression and decompression |
-
2018
- 2018-06-26 CN CN201880051817.8A patent/CN111357203A/zh active Pending
- 2018-06-26 US US16/019,513 patent/US10621108B2/en active Active
- 2018-06-26 KR KR1020207002460A patent/KR102366069B1/ko active Active
- 2018-06-26 JP JP2019572780A patent/JP6903773B2/ja active Active
- 2018-06-26 WO PCT/US2018/039620 patent/WO2019005888A1/en not_active Ceased
- 2018-06-26 KR KR1020217026017A patent/KR20210107136A/ko not_active Ceased
- 2018-06-26 EP EP18824155.8A patent/EP3646208A4/en not_active Withdrawn
- 2018-06-26 KR KR1020247003615A patent/KR102782120B1/ko active Active
-
2020
- 2020-03-18 US US16/823,160 patent/US11003589B2/en active Active
-
2021
- 2021-05-10 US US17/316,050 patent/US12596655B2/en active Active
- 2021-06-23 JP JP2021103991A patent/JP7488798B2/ja active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5838832A (en) * | 1995-04-03 | 1998-11-17 | Iterated Systems, Inc. | Method and system for representing a data set with a data transforming function and data mask |
| US20020154770A1 (en) * | 1999-11-09 | 2002-10-24 | Short Kevin M. | Method and apparatus for chaotic opportunistic lossless compression of data |
| US20150169462A1 (en) * | 2012-07-05 | 2015-06-18 | Blancco Oy Ltd | Apparatus, a system, a method and a computer program for erasing data stored on a storage device |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP3646208A4 * |
Also Published As
| Publication number | Publication date |
|---|---|
| JP2020527302A (ja) | 2020-09-03 |
| KR20200064054A (ko) | 2020-06-05 |
| KR102782120B1 (ko) | 2025-03-18 |
| JP2021166072A (ja) | 2021-10-14 |
| US11003589B2 (en) | 2021-05-11 |
| EP3646208A4 (en) | 2021-03-17 |
| EP3646208A1 (en) | 2020-05-06 |
| JP7488798B2 (ja) | 2024-05-22 |
| US12596655B2 (en) | 2026-04-07 |
| US10621108B2 (en) | 2020-04-14 |
| US20200218666A1 (en) | 2020-07-09 |
| KR20240019390A (ko) | 2024-02-14 |
| KR20210107136A (ko) | 2021-08-31 |
| US20190018793A1 (en) | 2019-01-17 |
| KR102366069B1 (ko) | 2022-02-23 |
| CN111357203A (zh) | 2020-06-30 |
| US20210326269A1 (en) | 2021-10-21 |
| JP6903773B2 (ja) | 2021-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US12596655B2 (en) | Systems and methods for transforming large data into a smaller representation and for re-transforming the smaller representation back to the original large data | |
| US11249999B2 (en) | Memory efficient searching | |
| US10572475B2 (en) | Leveraging columnar encoding for query operations | |
| US10853371B2 (en) | Efficient evaluation of query expressions including grouping clauses | |
| CN108292310B (zh) | 用于数字实体相关的技术 | |
| US11288275B2 (en) | Technique for fast join processing of dictionary encoded key columns in relational database systems | |
| CN111523001B (zh) | 用于存储数据的方法、装置、设备以及存储介质 | |
| US11074260B2 (en) | Space-efficient methodology for representing label information in large graph data for fast distributed graph query | |
| WO2016046233A1 (en) | Efficient conditional state mapping in a pattern matching automaton | |
| US9639549B2 (en) | Hybrid of proximity and identity similarity based deduplication in a data deduplication system | |
| US11403273B1 (en) | Optimizing hash table searching using bitmasks and linear probing | |
| Loncaric et al. | Pcodec: Better Compression for Numerical Sequences | |
| CN113728304A (zh) | 分布式存储系统中的粒度改变检测 | |
| US12229125B2 (en) | Selection pushdown in column stores using bit manipulation instructions | |
| US20250363074A1 (en) | System and Method for Arithmetic Operations on Compacted Data Files | |
| US12099475B2 (en) | System and method for random-access manipulation of compacted data files | |
| Zoellin et al. | New database compression assists in the IBM z14 processor | |
| US12578857B2 (en) | System and method for operating system multiloading and virtualization using codebook encoding of data streams and process calls | |
| Kumar et al. | Compendious and Optimized Succinct Data Structures for Big Data Store | |
| WO2025097078A2 (en) | System and method for operating system multiloading and virtualization using codebook encoding of data streams and process calls | |
| EP3036663B1 (en) | Thin database indexing | |
| CN115335823A (zh) | 用于最短路径图搜索的向量化的队列 |
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: 18824155 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2019572780 Country of ref document: JP Kind code of ref document: A |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| ENP | Entry into the national phase |
Ref document number: 20207002460 Country of ref document: KR Kind code of ref document: A |
|
| ENP | Entry into the national phase |
Ref document number: 2018824155 Country of ref document: EP Effective date: 20200127 |