Classifications

• • 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
  • G06F7/74—Selecting or encoding within a word the position of one or more bits having a specified value, e.g. most or least significant one or zero detection, priority encoders
• • 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
  • H03M7/40—Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code
  • H03M7/4031—Fixed length to variable length coding
  • H03M7/4037—Prefix coding
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
  • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
  • G06F3/0601—Interfaces specially adapted for storage systems
  • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
  • G06F3/0638—Organizing or formatting or addressing of data
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F18/00—Pattern recognition
  • G06F18/20—Analysing
  • G06F18/24—Classification techniques
  • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
  • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
  • G06F3/0601—Interfaces specially adapted for storage systems
  • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
  • G06F3/0608—Saving storage space on storage systems
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
  • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
  • G06F3/0601—Interfaces specially adapted for storage systems
  • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
  • G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
  • G06F3/0661—Format or protocol conversion arrangements
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
  • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
  • G06F3/0601—Interfaces specially adapted for storage systems
  • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
  • G06F3/0671—In-line storage system
  • G06F3/0673—Single storage device
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
  • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
  • G06F3/0601—Interfaces specially adapted for storage systems
  • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
  • G06F3/0671—In-line storage system
  • G06F3/0673—Single storage device
  • G06F3/0679—Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]
• • 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
  • G06F7/06—Arrangements for sorting, selecting, merging, or comparing data on individual record carriers
  • G06F7/08—Sorting, i.e. grouping record carriers in numerical or other ordered sequence according to the classification of at least some of the information they carry
• • 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
  • H03M7/40—Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code
  • H03M7/4031—Fixed length to variable length coding
  • H03M7/4037—Prefix coding
  • H03M7/4043—Adaptive prefix coding
  • H03M7/4056—Coding table selection
• • 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
  • H03M7/40—Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code
  • H03M7/4031—Fixed length to variable length coding
  • H03M7/4037—Prefix coding
  • H03M7/4043—Adaptive prefix coding
  • H03M7/4062—Coding table adaptation

Definitions

• a fingerprint is a short hash that encodes the presence of an item in the set that ASMDS approximately represents.
• fingerprints are stored in buckets (akin to cache sets) within the ASMDS, where the position of the fingerprint within the bucket does not change its meaning.
• Prior work orders these fingerprints by their prefix and replaces the prefixes with a code word. To encode and decode fingerprints, prior work uses a pair of encoding and decoding tables.
• FIG. 1 illustrates in block diagram form a data processing platform including compression according to some embodiments.
• FIG. 2 illustrates in flow diagram form a process for compressing and decompressing data according to some embodiments.
• FIG. 3 illustrates in flow diagram form a more detailed process for creating a code word containing compressed data according to some embodiments.
• FIG. 4 illustrates in table form a process for using encoding tables to create a code word containing compressed data according to some embodiments.
• FIG. 5 is a diagram showing a semi-sorting process according to some embodiments.
• FIG. 6 is a diagram showing an example of using encoding tables to create a code word according to some embodiments.
• FIG. 7 illustrates in flow diagram form a process for decompressing a code word according to some embodiments.
• FIG. 8 illustrates in diagram form part of a process using encoding tables to decompress a code word according to some embodiments.
• FIG. 9 illustrates in diagram form a further portion of the process of FIG. 8.
• FIG. 10 illustrates in chart form an example of compression performance obtained according to some exemplary embodiments.
• a data processing platform includes a memory and a processor coupled to the memory which is capable of performing compression and decompression of a set of data items.
• the processor receives a set of data items and selects suffix data and a prefix for each respective data item in the set of data items based on data content of the respective data item.
• the set of data items is sorted based on the prefixes.
• the prefixes are encoded by querying multiple encoding tables to create a code word containing compressed information representing values of all prefixes for the set of data items.
• the suffix data for each of the data items and the code word are stored in the memory.
• the code word is decompressed to recover the prefixes, and the recovered prefixes are paired with their respective suffix data.
• a method compresses and stores data items.
• a set of data items is received, and suffix data and a prefix are selected for each respective data item in the set of data items based on data content of the respective data item.
• the set of data items is sorted based on the prefixes.
• the prefixes are encoded by querying multiple encoding tables to create a code word containing compressed information representing values of all prefixes for the set of data items.
• the code word and the suffix data for each of the data items are stored.
• the code word is decompressed to recover the prefixes.
• the recovered prefixes are paired with their respective suffix data.
• a tangible nontransitory computer readable medium holds a program product executable by at least one processor to compress and decompress data items.
• a set of data items is received, and suffix data and a prefix are selected for each respective data item in the set of data items based on data content of the respective data item.
• the set of data items is sorted based on the prefixes.
• the prefixes are encoded by querying multiple encoding tables to create a code word containing compressed information representing values of all prefixes for the set of data items.
• the code word and the suffix data for each of the data items are stored in memory.
• the code word is decompressed to recover the prefixes.
• the recovered prefixes are paired with their respective suffix data.
• FIG. 1 illustrates in block diagram form a data processing platform including compression according to some embodiments.
• Data processing platform 100 may be any platform with a need for compressing a set of data items with high entropy that may be reordered.
• the techniques herein have many applications like compression of software and hardware cache tags and fingerprints in approximate set membership data structures.
• data processing platform 100 provides compression and decompression for a set of fixed-length items. The techniques are most beneficial with uniformly random items (i.e., with maximum informational entropy) whose order can be permuted. Application may vary from networking, software systems, and hardware caches.
• Data processing platform 100 communicates with a network or host system 10 to fulfill requests for data storage and retrieval, or perform operations such as encryption key retrieval or caching for a host system.
• Data processing platform 100 includes a processor 110 in communication with random-access memory (RAM) 112, which may be internal or external to processor 110.
• RAM 112 may be part of an integrated cache of processor 110 or in the platform 100’s main memory, for example.
• Processor 110 is also in communication with a non-volatile memory 120, which provides a tangible, non- transitory computer readable medium, such as Flash memory or a hard drive, that holds computer program code such as compression and decompression program code 122.
• RAM 112 holds data used in the compression techniques described below, such as a code word s, and suffix 0 to suffix k- ⁇ (there are a number k of suffixes, each of which is associated with one of the k data items). Such data may also be stored in non-volatile memory.
• program code 122 is loaded from non-volatile memory 120
• compression and decompression program code are present in RAM 112 as depicted by the loaded compression engine 114 and loaded decompression engine 116.
• Other data used with the techniques described herein, such as the data items to be compressed and encoding tables may also be held in RAM 112 or non-volatile memory 120.
• FIG. 2 illustrates in flow diagram form a process 200 for compressing and decompressing data according to some embodiments.
• a set of data items is provided for compression at block 202.
• the data items are typically fixed-length items, often have a high entropy as discussed above. Some subsequences of bits in the data items may have a high entropy, while other subsequences may be less random.
• various types of cache lines, cache tags, fingerprints from ASMDSs, network routing data, hardware tables from branch predictors or hardware prefetchers, or database entries to name some examples.
• process 200 selects a prefix and suffix data for each data item in the set based on the item’s data content.
• the data items may already have identified prefixes and suffixes, in which case block 204 is not necessary.
• the set of data items may be a set of cache lines with cache tags already provided, of which a portion or the full tag are used as a prefix.
• process 200 preferably selects a deterministic subsequence of bits from each item, such as a number of initial bits from the item.
• the subsequence may be selected from any desired part of the data item and may include non-sequential bits from the data item such as a fingerprint or hash that is computed from non-contiguous bit subsequences from the data item. Alternatively, a hash or computation on the desired part of the data item is used in certain implementations to select out or to produce a prefix.
• the suffix data is typically the bits remaining in the data item excluding the prefix.
• the data items are sorted based on the prefixes.
• the sorting may be done on the entire set of data items, or the set may be split into subsets and the subsets sorted. In some implementations, data items are stored in order subject to an ordering of the prefixes and for these embodiments, block 206 need not be executed.
• a code word representing compressed values of all the prefixes in the set is created, as further described below. The code word is created based on the sorted set. The code word may be non-lossy, that is, able to reproduce every bit of the compressed prefixes. If multiple subsets are sorted, a code word is typically created for each subset.
• the suffix data and the code word are stored in memory.
• the memory used for storage may be short term RAM such as RAM 112 (FIG. 1) or storage memory such as non-volatile memory 120 (FIG. 1).
• the suffix data and code word may also be stored in a database or networked storage that is not part of the data processing platform performing the compression and decompression.
• the stored suffix data and code word are retrieved, and the prefixes are decompressed at block 212.
• An example of one suitable decompression process is described below.
• the decompressed prefixes are then paired with the prefix data to provide the complete data items.
• the complete code word does not need to be decompressed, if the desired data items are extracted early in the decompression process, as further described below.
• FIG. 3 illustrates in flow diagram form a more detailed process 300 for creating a code word containing compressed data according to some embodiments. Further examples of compression processes are illustrated in FIGS. 4-6. These processes may be performed by a software compression engine such as compression engine 114 of FIG. 1, or by other suitable combinations of hardware and software.
• a software compression engine such as compression engine 114 of FIG. 1, or by other suitable combinations of hardware and software.
• Process 300 begins creating a code word at block 302.
• Block 304 provides the encoding tables needed to create the code word by either retrieving them from memory, calculating them, or partially calculating them, as further described below.
• block 306 For each prefix, in the order provided by sorting, block 306 encodes the prefixes by querying multiple encoding tables to create a code word s containing compressed information representing values of all prefixes for the set of data items. In this embodiment, block 306 queries the encoding tables to obtain a plurality of integers representing respective prefix positions in the encoding tables.
• the process may work with variation in the sorting results, as long as the data items are ordered via their prefix values.
• the results of querying the multiple input tables are used to create the code word s at block 308, in this embodiment by summing the resulting integers as shown.
• FIG. 5 shows the raw input of a set of data items with the prefixes and suffixes identified.
• the depicted data items are merely one example with a small number of small data items for simplicity, and many other arrangements of data items may be stored according to the techniques herein.
• These data items are sorted as set forth at block 206 (FIG. 2).
• This sort constitutes a semi-sort with respect to the data items because the suffix data is not used in the sort, only the prefix data.
• the results of the sort may therefore vary in order, as shown by the sort output A and B of FIG. 5, which are both valid results of a prefix-based sort on the data items. Notice that the two data items that share the same prefix (101) appear in swapped positions in A and B. More generally, data items with the same prefix may appear in any order in the output of the sort among one another without affecting the output’s correctness.
• Process 300 is effective despite possible variations in the order of the input data such as those in FIG. 5.
• FIG. 4 illustrates in table form a process for using encoding tables to create a code word containing compressed data according to some embodiments.
• the compression encoding takes an ordered list of k , r-bit prefixes and outputs a codeword that expresses them.
• Integer k represents the number of data items in the set
• r represents the number of bits in each prefix.
• One such encoding scheme is The scheme uses an iterator i that enumerates through the data items in a set 0, 1 , 2, ... , k — 2 , k — 1.
• This encoding can be done by accessing one or more precomputed tables, by computing the closed form expressions in response to a compression or decompression request requiring the computed values, or by a combination thereof.
• Yet another alternative is to only compute values or partial values as they are needed but cache the values after they are computed in an associated data or hardware structure for later use. It is also possible to reduce the multiplication calculations involved in the computation down to a series of equivalent additions, shift, and masking operations.
• the number of tables is equal to the number of prefixes k. In other embodiments, fewer tables may be used. For example, it is possible to use implicit encoding to implement table / ⁇ as further discussed below with respect to FIG. 10.
• the tables are generally referred to as tables l 0 through l k® , where each l t is the encoding table for the ith data item’s prefix p t . Each l t holds 2 r entries and is indexed into by the specific j , which yields the specific encoding.
• index values representing the prefix position within the table are used to encode the prefixes.
• other values may be used.
• the indexed values may appear in a different order in the table from that depicted, in which case the index returned is not necessarily a prefix position.
• Various techniques may be used to increase the efficiency of computing or storing the encoding tables.
• the encoding tables could be partially computed on the fly to reduce their storage cost. For instance, every multiply by 2 may be implemented as a shift operation, so it is more efficient to store a count of such shifts rather than the shifted values.
• the process may replace the individual /, s with a single matrix that is indexed by the tuple p l and i, which does not store duplicate values for when i + 1 is equal to p t - 1 (e.g., the process does not store both Q and
• the encoding tables may be no more than a kilobyte in size for common values of r and k.
• the process need only store on the order of 2 r k entries, so values of r in the range of about 4 to 8 may be employed while still providing the majority of the space savings and with a small area overhead for the encoding tables.
• the reduction in storage requirements for an exemplary embodiment in this range of r values can be seen below in FIG. 10.
• FIG. 5 illustrates in diagram form results 500 of a semi-sorting process according to some embodiments. Depicted are four raw input data items to the process, each with 8 bits. The 3-bit prefixes and the remaining 5-bit suffixes are identified. Two valid variations of the semi- sorted output (the output of block 206, FIG. 2, for example) are shown labelled A and B. As can be seen, two data items in the set have a prefix value of 101.
• FIG. 6 illustrates in diagram form an exemplary process 600 of using encoding tables to create a code word according to some embodiments.
• the depicted scenario uses the set of data items from FIG. 5, sorted according to output A.
• the encoding process generally starts with of k , r- bit prefixes, and encodes the prefixes to produce the code word s that is the compressed representation of the prefixes.
• k is four
• process 600 uses four encoding tables h to k.
• the encoding tables may be produced according to the techniques described with respect to FIG. 4, or with other suitable techniques.
• the diagram shows an arrow from each prefix value to a table entry in the respective encoding table for the prefix, indicating the prefix value is used as an index into the table.
• the function “nCf ’ indicates the combinatorial expression " n choose k ,” the number of ways to choose k items from a group of n items if order does not matter.
• An integer value results at each respective entry, as described with respect to FIG. 3, block 306.
• the arrows out of the table entry to the adder block represent that the integer values are summed to produce the code word s, in this example having a value of 261. This code word value is stored in memory along with the 5-bit suffix values.
• the encoding tables are provided such that, for each prefix, the respective prefix index provides a count based on a plurality of ordered collections of prefix values ordered by an ordering property, with the counts based on the prefix’s relative position in the ordered collections.
• the count for each respective prefix index indicates a number of ordered collections of prefix values where each ordered collection’s size is the output of a function applied to the prefix values’ relative position in the ordered collections, and all prefix values in the ordered collections are less than the value of the respective prefix as computed by the ordering property. For example, in the embodiment of FIG.
• the zth prefix when examining the zth prefix, where 0 ⁇ i ⁇ k, the zth prefix’s encoding is retrieved by performing a table lookup on /, [prefix].
• the index value returned is a count of the number of length i I sorted lists of prefixes that numerically precede the first list containing zth prefix. For instance, if z is 3 and the prefix is 2, /, 121 stores the count of the number of length 3+1 lists ordered in nonincreasing order that precede the list 2, 0, 0, 0,
• FIG. 7 illustrates in flow diagram form a process 700 for decompressing a code word according to some embodiments.
• FIG. 8 illustrates in diagram form part of a process 800 using encoding tables to decompress a code word according to some embodiments.
• FIG. 9 is a diagram showing a further portion 900 of the process of FIG. 8. The following description will refer to FIGS. 7-9 as needed.
• Process 700 may be performed by a software decompression engine 116 (FIG. 1), or other suitable hardware implementation.
• the depicted decoding process is one example that may be used given the encoding process described above. Others decoding schemes are possible for the encoding used above or other encodings.
• process 700 retrieves the code word s for the set of data items from memory and begins decompression.
• the stored suffixes of the set of data items are retrieved as well, as shown along the top of FIG. 8.
• the decoding generally starts with the code word s that is the compressed representation of k , r-bit prefixes, and decodes the prefixes starting from the A>lth prefix down to the 0th prefix.
• Process 700 incudes k iterations, in this example four, one per prefix.
• the iterations employ encoding tables k to k, as shown in FIGS. 8 and 9, in the order produced by the semi-sorting.
• the tables used are the same encoding tables that were employed in encoding from table h-i to table l 0 .
• process 700 at block 704 initializes an integer i to be k - I and a search value s' to be equal to the code word s. Then at block 706, searching encoding table /, for a largest value that is less than or equal to the search value s' . This is illustrated in FIG. 8 where table k is shown with the search result highlighted from the searching for the highest value less than s' of 261, which is the value 210 found in the last entry.
• This search result is used at block 708, which sets the recovered prefix p, equal to a table index of the largest value found in the search.
• the recovered prefix is set to the index 111 of the last entry in table 13.
• the recovered prefix value of 111 is combined with the stored suffix value 10111 , in the order provided by the sort used when encoding the code word, providing the complete data item shown under table h.
• the integer / is set decremented to proceed to the next iteration.
• the encoding table / is searched for the largest value that is less than or equal to the new search value s' . This is shown in FIG. 8 by the highlighted search result in table h of 35, the largest table entry less than 51.
• the recovered prefix p is set equal to a table index of the largest value found in the search. In the example recovered prefix p: is set to the highlighted index value 101 for table h. This prefix is combined with the stored suffix data to provide the complete data item for prefix p-. which is 10111111.
• Blocks 710 through 716 are repeated until a final iteration is completed with integer / at zero, which completes the set of recovered prefixes.
• the final two iterations are depicted showing resulting values.
• Table k is searched with a value of s[ of 16, which equals the prior search value 51 from the prior iteration minus the largest value found, 35.
• the recovered index pi is set to the table index value found searching with 16, which is 101, the index of 15 from table k.
• table h is searched with a value of s ⁇ of 1, which equals the prior search value 16 minus the largest value previously found, 15.
• the recovered index po is set to the table index value found searching with 1, which is 001, the index of the value 1 found in table lo.
• the searches of blocks 706 and 714 may be done in a linear or parallel fashion for small tables or alternatively using another suitable search algorithm such as a modified binary search for larger ones.
• the process may be implemented with a microprocessor executing program code, with application specific digital logic, or with programmable logic.
• FIG. 10 illustrates in chart form an example of compression performance obtained according to some embodiments.
• Chart 1000 shows performance of the example process described above for multiple data sets, with the reduction in bits per data item on the vertical axis versus the prefix length in bits (r), the number of bits compressed per item, on the horizontal axis.
• the chart legend identifies the different associativities of the data sets.
• the term “associativity” is used herein as it is generally used in the art to describe cache sets or similar sets of data. That is, each memory address or other data item identifier maps to a possible number of positions within a set or group of data items. The number of possible positions is the associativity. Note that each power of 2 increase in associativity yields approximately another bit in savings per compressed prefix (assuming an appropriate value for r).
• the initial space to store those prefixes is kr bits.
• Such an encoding allows the storage process to encode all 2 rk permutations of a set of k , r-bit numbers where duplicates are permissible.
• the problem evolves from having to store all permutations of k r- bit numbers to only having to store the combinations.
• the techniques herein can encode those combinations in bits, when rounding up to the nearest whole bit.
• the net savings per set are therefore the net size of k , r-bit numbers (kr bits) minus the new size (f log 2 2 + fc fc _1 )
• the storage cost of the prefixes can often be reduced by one to four bits depending on the set associativity.
• ASMDSs where data items are often only eight to a few tens of bits in length, the savings of one to four bits per item are significant.
• the prefixes only need to be several bits in length to realize the bulk of the benefit of compression. This property is important because the individual encoding lookup tables in the example processes herein each have 2 r entries, so a longer prefix is costly in terms of the encoding table size.
• the storage cost of this approach scales roughly linearly with k when r is fixed but diverges from linear behavior because each lookup table entry takes on the order of log2 ⁇ 2 + fc fc_1 ) bits to encode. Performing the encoding and decoding by computing the lookup table values on the fly further reduces the storage costs.
• the final lookup table implicit because it is always ( ⁇ ), which is simply p 0 , so whatever is remaining after k - 1 iterations of the decompression process is the last prefix.
• the final prefix of the encoding process may be encoded into the code word by adding the zeroth prefix to the integer sum (FIG.
• the techniques described herein may be implemented with various combinations of hardware and software.
• the hardware circuitry may include digital logic, finite state machines, programmable logic arrays (PLAs), and the like.
• the compression and decompression processes could be implemented with a microcontroller executing stored program instructions to evaluate the relative timing eligibility of the pending commands.
• some of the instructions may be stored in a non-transitory computer memory or computer readable storage medium for execution by the microcontroller.
• the non-transitory computer readable storage medium includes a magnetic or optical disk storage device, solid-state storage devices such as Flash memory, or other non-volatile memory device or devices.
• the computer readable instructions stored on the non-transitory computer readable storage medium may be in source code, assembly language code, object code, or other instruction format that is interpreted and/or executable by one or more processors.
• the data processing platform 100 of FIG. 1, or any portions thereof, such as processor 110, or other implementations such as customized digital logic implementations, may be described or represented by a computer accessible data structure in the form of a database, database index, or other data structure which can be read by a program and used, directly or indirectly, to fabricate integrated circuits.
• this data structure may be a behavioral-level description or register-transfer level (RTL) description of the hardware functionality in a high-level design language (HDL) such as Verilog or VHDL.
• HDL high-level design language
• VHDL Verilog
• the netlist includes a set of gates that also represent the functionality of the hardware including integrated circuits.
• the netlist may then be placed and routed to produce a dataset describing geometric shapes to be applied to masks.
• the masks may then be used in various semiconductor fabrication steps to produce the integrated circuits.
• the database on the computer accessible storage medium may be the netlist (with or without the synthesis library) or the data set, as desired, or Graphic Data System (GDS) II data.
• GDS Graphic Data System
• the type of data in the data sets may vary in different embodiments, such as various types of cache lines, cache tags, fingerprints from ASMDSs, network routing data, hardware tables from branch predictors or hardware prefetchers, or database entries.
• the mathematical algorithm used to produce the encoding tables may vary.
• the particular process for encoding and decoding based on the multiple tables may also vary.
• the method of providing the encoding tables may also vary, including calculation or partial calculation of the table values.
• a single code word is employed in the examples herein, other embodiments may produce multiple code words for a particular data set, depending on the compression algorithm employed.

Landscapes

• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Engineering & Computer Science (AREA)
• General Physics & Mathematics (AREA)
• Human Computer Interaction (AREA)
• Data Mining & Analysis (AREA)
• Life Sciences & Earth Sciences (AREA)
• Artificial Intelligence (AREA)
• Bioinformatics & Cheminformatics (AREA)
• Bioinformatics & Computational Biology (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Evolutionary Biology (AREA)
• Evolutionary Computation (AREA)
• Compression, Expansion, Code Conversion, And Decoders (AREA)

Priority Applications (4)

Applications Claiming Priority (2)

Publications (1)

Family

ID=74567451

Family Applications (1)

Country Status (6)

Families Citing this family (2)

Citations (5)

Family Cites Families (17)

• 2019
  • 2019-08-16 US US16/542,872 patent/US11309911B2/en active Active
• 2020
  • 2020-08-12 CN CN202080058010.4A patent/CN114258521B/zh active Active
  • 2020-08-12 KR KR1020227007999A patent/KR102824624B1/ko active Active
  • 2020-08-12 EP EP20854389.2A patent/EP4014128A4/en active Pending
  • 2020-08-12 JP JP2022508996A patent/JP7631308B2/ja active Active
  • 2020-08-12 WO PCT/US2020/045903 patent/WO2021034565A1/en not_active Ceased
• 2022
  • 2022-04-18 US US17/722,931 patent/US11736119B2/en active Active

Patent Citations (5)

Non-Patent Citations (2)

Also Published As

Similar Documents

Legal Events