Abstract: Technologies for refining stochastic similarity search candidates include a device having a memory that is column addressable and circuitry connected to the memory. The circuitry is configured to add a set of input data vectors to the memory as a set of binary dimensionally expanded vectors, including multiplying each input data vector with a projection matrix. The circuitry is also configured to produce a search hash code from a search data vector, including multiplying the search data vector with the projection matrix. Additionally, the circuitry is configured to identify a result set of the binary dimensionally expanded vectors as a function of a Hamming distance of each binary dimensionally expanded vector from the search hash code and determine, from the result set, a refined result set as a function of a similarity measure in an original input space of the input data vectors.

Abstract: Methods, apparatus, systems, and articles of manufacture are disclosed to facilitate improved use of stochastic associative memory. Example instructions cause at least one processor to: generate a hash code for data to be stored in a stochastic associative memory (SAM); compare the hash code with centroids of clusters of data stored in the SAM; select a first one of the clusters corresponding to a first one of the centroids that is closest to the hash code; determine whether a selected number of hash codes stored in the SAM exceeds a threshold; in response to the selected number exceeding the threshold: query a controller for sizes of the clusters; and determine, based on the query, that a second one of the clusters includes an unbalanced size; and select a third one of the clusters to associate with a second number of hash codes corresponding to the second one of the clusters.

Abstract: Technologies for performing stochastic similarity searches in an online clustering space include a device having a column addressable memory and circuitry. The circuitry is configured to determine a Hamming distance from a binary dimensionally expanded vector to each cluster of a set of clusters of binary dimensionally expanded vectors in the memory, identify the cluster having the smallest Hamming distance from the binary dimensionally expanded vector, determine whether the identified cluster satisfies a target size, and add or delete, in response to a determination that the identified cluster does not satisfy the target size, the binary dimensionally expanded vector to or from the identified cluster.

Abstract: Technologies for tuning performance and/or accuracy of similarity search using stochastic associative memories (SAM). Under a first subsampling approach, columns associated with set bits in a search key comprising a binary bit vector are subsampled. Matching set bits for the subsampled columns are aggregated on a row-wise basis to generate similarity scores, which are then ranked. A similar scheme is applied for all the columns with set bits in the search key and the results for top ranked rows are compared to evaluate a tradeoff between throughput boost versus lost accuracy. A second approach called continuous column read, and iterative approach is employed that continuously scores the rows as each new column read is complete. The similarity scores for an N-1 and Nth-1 iteration are ranked, a rank correlation is calculated, and a determination is made to whether the rank correlation meets or exceeds a threshold.

Abstract: An example of an apparatus may include memory organized as at least one bank that includes two or more arrays, and circuitry communicatively coupled to the memory to select respective rows of the two or more arrays of a bank for a memory access operation based on an access orientation signal. Other examples are disclosed and claimed.

Abstract: Column read enabled three dimensional cross-point memory is optimized to reduce delay time due to partition busy times incurred when reading from a same partition. A column read enabled memory media stores each entry of a logical column of an array of bits in contiguous different physical rows and different physical columns of the cross-point memory array than any other entry of the logical column. Subsets of the contiguous different physical rows are stored in different partitions to reduce the delay time incurred when performing column reads from a same partition to improve media management operation performance.

Abstract: Technologies for column reads for clustered data include a device having a column-addressable memory and circuitry connected to the memory. The column-addressable memory includes multiple dies. The circuitry may be configured to determine multiple die offsets based on a logical column number of the data cluster, determine a base address based on the logical column number, program the dies with the die offsets. The circuitry is further to read logical column data from the column-addressable memory. To read the data, each die adds the corresponding die offset to the base address. The column-addressable memory may include multiple command/address buses. The circuitry may determine a starting address for each of multiple logical columns and issue a column read for each starting address via a corresponding command/address bus. Other embodiments are described and claimed.

Abstract: Technologies for media management for providing column data layouts for clustered data include a device having a column-addressable memory and circuitry connected to the memory. The circuitry is configured to store a data cluster of a logical matrix in the column-addressable memory with a column-based format and to read a logical column of the data cluster from the column-addressable memory with a column read operation. Reading the logical column may include reading logical column data diagonally from the column-address memory, including reading from the data cluster and a duplicate copy of the data cluster. Reading the logical column may include reading from multiple complementary logical columns. Reading the logical column may include reading logical column data diagonally with a modulo counter. The column data may bread from a partition of the column-address memory selected based on the logical column number. Other embodiments are described and claimed.

Abstract: A memory accessed by rows and/or by columns in which an array of bits can be physically stored physical one-bit wide columns with each bit of the multi-bit wide logical column stored in a one-bit physical column in a different physical die. The multi-bit column is read by reading a one-bit physical column in each of the different physical die in parallel. The multi-bit wide logical column is arranged diagonally across M physical rows and M one-bit physical columns with each bit of the multi-bit wide logical column in the logical row stored in a different physical row and physical one-bit wide column in one of plurality of dies.

Abstract: Technologies for performing encoding of data symbols for column read operations include a device having a memory that is column addressable and circuitry connected to the memory. The circuitry is configured to obtain a data set to encode. The data set is defined by a set of data symbols. The circuitry is also configured to determine a set of codewords to encode the data symbols of the data set, including defining each codeword with a set bit distance of at least two from every other codeword in the set of codewords. Additionally, the circuitry is configured to write the data set to the memory as a function of the determined set of codewords.

Abstract: A memory accessed by rows and/or by columns in which an array of bits can be physically stored in multi-bit wide columns in physically contiguous rows is provided. A multi-bit wide logical column is arranged diagonally across (M/multi-bits) physical rows and (M/multi-bits) physical columns with each of the plurality of multi-bit wide logical columns in the logical row stored in a different physical row and physical multi-bit column.

Abstract: Binary sparse encoding of data can be used to reduce an amount of data read from the stochastic associative memory while processing a query. Read performance of the stochastic associated memory is optimized to enhance the query throughput by modifying access patterns to reduce the time to read the stochastic associated memory. Read performance of the stochastic associative memory can be further improved through the use of cluster aware sharding and replication for parallelized similarity search. Clusters are partitioned across multiple Dual In-line Memory Modules (DIMMs), each DIMM including stochastic associative memory, to achieve maximum latency advantage.

Abstract: Methods and apparatus for sparse column-aware encodings for numeric data types, including integer data and floating-point data (float, double, etc.). The encoding schemes are tailored to take advantage of column addressable memories such as stochastic associative memories (SAM) to enable Stochastic Associative Search (SAS), which is a highly efficient and fast way of searching through a very large database of records (order of Billions) and finding similar records to a given query record (search key). Techniques are also disclosed for performing range searches for both integer and floating-point data types. The integer or float data is converted to Hexadecimal form and encoded using an m-of-n constant weight encoding. Only the columns with set bits in search keys need to be read, which significantly reduces the number of reads required for searches.

Abstract: Systems, apparatuses and methods may provide for technology that organizes data and corresponding parity information into a plurality of die words, distributes a column of the die words across a plurality of storage dies, and distributes the column across a plurality of partitions. In one example, the technology also reads a row of the die words at a read rate and reads the column of the die words at the read rate.

Abstract: Examples described herein relate to an apparatus comprising: circuitry to receive a request to store data as a part of a matrix in a memory device; circuitry to allocate address mappings to the data to reduce a number of sequential accesses to a same partition of a portion of the memory device; circuitry to store the address mappings for access with a read operation; and circuitry to cause storage of the data into the memory device according to the address mappings. In some examples, the matrix comprises one or more columns and/or one or more rows. In some examples, the memory device comprises one or more of: a three-dimensional (3D) cross point memory device, volatile memory device, or non-volatile memory device.

Abstract: Technologies for tuning performance and/or accuracy of similarity search using stochastic associative memories (SAM). Under a first subsampling approach, columns associated with set bits in a search key comprising a binary bit vector are subsampled. Matching set bits for the subsampled columns are aggregated on a row-wise basis to generate similarity scores, which are then ranked. A similar scheme is applied for all the columns with set bits in the search key and the results for top ranked rows are compared to evaluate a tradeoff between throughput boost versus lost accuracy. A second approach called continuous column read, and iterative approach is employed that continuously scores the rows as each new column read is complete. The similarity scores for an N-1 and Nth-1 iteration are ranked, a rank correlation is calculated, and a determination is made to whether the rank correlation meets or exceeds a threshold.

Abstract: Technologies for media management for providing column data layouts for clustered data include a device having a column-addressable memory and circuitry connected to the memory. The circuitry is configured to store a data cluster of a logical matrix in the column-addressable memory with a column-based format and to read a logical column of the data cluster from the column-addressable memory with a column read operation. Reading the logical column may include reading logical column data diagonally from the column-address memory, including reading from the data cluster and a duplicate copy of the data cluster. Reading the logical column may include reading from multiple complementary logical columns. Reading the logical column may include reading logical column data diagonally with a modulo counter. The column data may bread from a partition of the column-address memory selected based on the logical column number. Other embodiments are described and claimed.

Abstract: Technologies for column reads for clustered data include a device having a column-addressable memory and circuitry connected to the memory. The column-addressable memory includes multiple dies. The circuitry may be configured to determine multiple die offsets based on a logical column number of the data cluster, determine a base address based on the logical column number, program the dies with the die offsets. The circuitry is further to read logical column data from the column-addressable memory. To read the data, each die adds the corresponding die offset to the base address. The column-addressable memory may include multiple command/address buses. The circuitry may determine a starting address for each of multiple logical columns and issue a column read for each starting address via a corresponding command/address bus. Other embodiments are described and claimed.

Abstract: Technologies for performing stochastic similarity searches in an online clustering space include a device having a column addressable memory and circuitry. The circuitry is configured to determine a Hamming distance from a binary dimensionally expanded vector to each cluster of a set of clusters of binary dimensionally expanded vectors in the memory, identify the cluster having the smallest Hamming distance from the binary dimensionally expanded vector, determine whether the identified cluster satisfies a target size, and add or delete, in response to a determination that the identified cluster does not satisfy the target size, the binary dimensionally expanded vector to or from the identified cluster.

Abstract: Technologies for performing random sparse lifting and Procrustean orthogonal sparse hashing using column read-enabled memory include a device that has a memory that is column addressable and circuitry connected to the memory. The circuitry is configured to add a set of input data vectors to the memory as a set of binary dimensionally expanded vectors, including multiplying each input data vector with a projection matrix. The circuitry is also configured to produce a search hash code from a search data vector, including multiplying the search data vector with the projection matrix. Further, the circuitry is configured to determine a Hamming distance between the search hash code and each of the binary dimensionally expanded vectors.