Abstract: A method and system for reducing computation of matrices in an application such as a large language model in ciphertext format and plaintext format. The example method uses singular value decomposition to zero out values in a matrix such as a weight matrix to produce a reduced matrix. The reduced matrix is used to determine two smaller matrices, which are multiplied with each other and the reduced weight matrix. Since multiplication and memory access are performed on smaller matrices, computational resources required are less in the example method. The example method thus reduces the complexity of matrix based calculations.

Abstract: A memory includes a read/write amplifier configured to retrieve read data from a memory array, and a neural network based preconditioning circuit configured to receive a read data signal according to the read data. A neural network of the preconditioning circuit is configured to precondition the read data signal based on a characteristic of a read data transmission path to provide a modified read data signal. The memory further includes an output driver configured to transmit the modified read data signal.

Abstract: Examples described herein include systems and methods which include wireless devices and systems with examples of mixing input data with coefficient data specific to a processing mode selection. For example, a computing system with processing units may mix the input data for a transmission in a radio frequency (RF) wireless domain with the coefficient data to generate output data that is representative of the transmission being processed according to a specific processing mode selection. The input data is mixed with coefficient data at layers of multiplication/accumulation processing units (MAC units). The processing mode selection may be associated with an aspect of a wireless protocol. Examples of systems and methods described herein may facilitate the processing of data for 5G wireless communications in a power-efficient and time-efficient manner.

Abstract: Apparatuses and methods for error correction based on data characteristics are disclosed. Data characteristics can include importance of the data. Data is received at a memory controller from a host device, and a characteristic of the received data is determined. A level of error correction is selected from a plurality of error correction levels for the received data based on the determined characteristic. The received data and an error correction code are written to a memory. The error correction code is generated based on the selected level of error correction. In some implementations, the characteristic of the received data is determined using a neural network.

Abstract: A system and method to reduce fast Fourier transforms (FFT) required for bootstrapping in a Fully Homomorphic Encryption process. Ciphertext is separated into a vector of n samples. A fast Fourier transfer (FFT) is performed over a first vector of the samples and a FFT over a bootstrap key. A phase vector of n W(a) terms that are equivalents to the output of a FFT is calculated. A first set of pointwise multiplications of the FFT of the first vector of samples with each of the n W(a) terms in the phase vector is performed. A second set of pointwise multiplications of the results of the first set of pointwise multiplications with the FFT of the bootstrap key is performed. An inverse FFT(IFFT) on the accumulated result of the second set of point-wise multiplications is performed to obtain a bootstrapping result of the ciphertext.

Abstract: A system includes a first wireless communication device comprising a first baseband processor neural network configured to process at least part of data for transmission to a second wireless communication device according to a collaborative processing configuration while collaborative processing is enabled to generate a first radio frequency (RF) signal. The first wireless communication device is configured to transmit the first RF signal. The system further includes a third wireless communication device comprising a second baseband processor neural network configured to, while the collaborative processing is enabled, process at least part of the data for transmission to the second wireless communication device according to a collaborative processing configuration to generate a second RF signal. The third wireless communication device is configured to transmit the second RF signal in collaboration with transmission of the first RF signal by the first baseband processor.

Abstract: Apparatuses, systems, and methods related to an image processor formed in an array of memory cells are described. An image processor as described herein is configured to reduce complexity and power consumption and/or increase data access bandwidth by performing image processing in the array of memory cells relative to image processing by a host processor external to the memory array. For instance, one apparatus described herein includes sensor circuitry configured to provide an input vector, as a plurality of bits that corresponds to a plurality of color components for an image pixel, and an image processor formed in an array of memory cells. The image processor is coupled to the sensor circuitry to receive the plurality of bits of the input vector. The image processor is configured to perform a color correction operation in the array by performing matrix multiplication on the input vector and a parameter matrix to determine an output vector that is color corrected.

Abstract: Apparatuses and methods for error correction based on data characteristics are disclosed. Data characteristics can include importance of the data. Data is received at a memory controller from a host device, and a characteristic of the received data is determined. A level of error correction is selected from a plurality of error correction levels for the received data based on the determined characteristic. The received data and an error correction code are written to a memory. The error correction code is generated based on the selected level of error correction. In some implementations, the characteristic of the received data is determined using a neural network.

Abstract: The disclosure is directed toward a method and system to efficiently determine an attention function by reducing bootstrapping steps and ciphertext matrix operations. A combined key and query matrix in plaintext is pre-calculated. A ciphertext-plaintext matrix multiplication of the combined key and query matrix with a ciphertext input is performed. The output of the resulting ciphertext-plaintext matrix multiplication of the combined key and query matrix is bootstrapped. A ciphertext-plaintext matrix multiplication of the value matrix with the ciphertext input is performed. The resulting ciphertext-plaintext matrix multiplication of the value matrix is bootstrapped.

Abstract: The disclosure is directed toward a method to efficiently perform a large language model operation by reducing the memory and computation resources for operating on a matrix such as a weight matrix. The weight matrix may be included in operations such as the head attention function of the large language model. The weight matrix is decomposed into multiple submatrices that each include a set of weights from the weight matrix. An input of the large language model is provided. A submatrix solution for the submatrix is determined from the input and the set of weights of the submatrix. The determining of a solution is repeated for each of the submatrices to provide multiple submatrix solutions. The resulting submatrix solutions are combined to obtain a matrix solution for the weight matrix.

Abstract: A system includes a first wireless communication device comprising a first baseband processor neural network configured to process at least part of data for transmission to a second wireless communication device according to a collaborative processing configuration while collaborative processing is enabled to generate a first radio frequency (RF) signal. The first wireless communication device is configured to transmit the first RF signal. The system further includes a third wireless communication device comprising a second baseband processor neural network configured to, while the collaborative processing is enabled, process at least part of the data for transmission to the second wireless communication device according to a collaborative processing configuration to generate a second RF signal. The third wireless communication device is configured to transmit the second RF signal in collaboration with transmission of the first RF signal by the first baseband processor.

Abstract: Apparatuses, systems, and methods related to an image processor formed in an array of memory cells are described. An image processor as described herein is configured to reduce complexity and power consumption and/or increase data access bandwidth by performing image processing in the array of memory cells relative to image processing by a host processor external to the memory array. For instance, one apparatus described herein includes sensor circuitry configured to provide an input vector, as a plurality of bits that corresponds to a plurality of color components for an image pixel, and an image processor formed in an array of memory cells. The image processor is coupled to the sensor circuitry to receive the plurality of bits of the input vector. The image processor is configured to perform a color correction operation in the array by performing matrix multiplication on the input vector and a parameter matrix to determine an output vector that is color corrected.

Abstract: The present invention relates to a memory controller and a memory device that are configured to communicate with each other using multiple input multiple output (MIMO) technology. The memory controller includes a precoder that precodes data for transmission. The precoding is based on channel state information, a neural network, or both. The memory device receives the precoded data and decodes them to retrieve the original data. In some cases, the precoder uses the channel state information to optimize the precoding matrix for the given channel conditions. In some cases, a neural network is trained to predict the optimal precoding matrix for the current channel state. The precoding matrix is then used to encode the data, which is then transmitted to the memory device. The use of MIMO and precoding improves the reliability and efficiency of the communication between the memory controller and memory device.

Abstract: Examples described herein include systems and methods which include wireless devices and systems with examples of configuration modes for baseband units (BBU) and remote radio heads (RRH). For example, a computing system including a BBU and a RRH may receive a configuration mode selection including information indicative of a configuration mode for respective processing units of the BBU and the RRH. The computing system may allocate the respective processing units to perform wireless processing stages associated with a wireless protocol. The BBU and/or the RRH may generate an output data stream based on the mixing of coefficient data with input data at the BBU and/or the RRH. Examples of systems and methods described herein may facilitate the processing of data for 5G (e.g., New Radio (NR)) wireless communications in a power-efficient and time-efficient manner.

Abstract: Examples described herein include systems and methods which include wireless devices and systems with examples of multiple frequency bands transmission with a recurrent neural network that compensates for the self-interference noise generated by power amplifiers at harmonic frequencies of a respective wireless receiver. The recurrent neural network may be coupled to antennas of a wireless device and configured to generate the adjusted signals that compensate self-interference. The recurrent neural network nay include a network of processing elements configured to combine transmission signals into sets of intermediate results. Each set of intermediate results may be summed in the recurrent neural network to generate a corresponding adjusted signal. The adjusted signal is receivable by a corresponding wireless receiver to compensate for the self-interference noise generated by a wireless transmitter transmitting on the same or different frequency band as the wireless receiver is receiving.

Abstract: The present invention relates to a memory controller and a memory device that are configured to communicate with each other using multiple input multiple output (MIMO) technology. The memory controller includes a precoder that precodes data for transmission. The precoding is based on channel state information, a neural network, or both. The memory device receives the precoded data and decodes them to retrieve the original data. In some cases, the precoder uses the channel state information to optimize the precoding matrix for the given channel conditions. In some cases, a neural network is trained to predict the optimal precoding matrix for the current channel state. The precoding matrix is then used to encode the data, which is then transmitted to the memory device. The use of MIMO and precoding improves the reliability and efficiency of the communication between the memory controller and memory device.

Abstract: Examples described herein include systems and methods which include wireless devices and systems with examples of full duplex compensation with a self-interference noise calculator. The self-interference noise calculator may be coupled to antennas of a wireless device and configured to generate adjusted signals that compensate self-interference. The self-interference noise calculator may include a network of processing elements configured to combine transmission signals into sets of intermediate results. Each set of intermediate results may be summed in the self-interference noise calculator to generate a corresponding adjusted signal. The adjusted signal is received by a corresponding wireless receiver to compensate for the self-interference noise generated by a wireless transmitter transmitting on the same frequency band as the wireless receiver is receiving.

Abstract: Methods and apparatus for performing diversity matrix operations within a memory fabric and for converting a memory array into a matrix fabric for spatial diversity-related matrix transformations and performing matrix operations therein. Exemplary embodiments described herein perform MIMO-related matrix transformations (e.g., precoding, beamforming, or data recovery matrix operations) within a memory device that includes a matrix fabric and matrix multiplication unit (MMU). In one variant, the matrix fabric uses a “crossbar” construction of resistive elements. Each resistive element stores a level of impedance that represents the corresponding matrix coefficient value. The crossbar connectivity can be driven with an electrical signal representing the input vector as an analog voltage. The resulting signals can be converted from analog voltages to a digital values by an MMU to yield a matrix-vector product. The MMU may additionally perform various other logical operations within the digital domain.

Abstract: Examples described herein include systems and methods which include wireless devices and systems with examples of full duplex compensation with a self-interference noise calculator. The self-interference noise calculator may be coupled to antennas of a wireless device and configured to generate adjusted signals that compensate self-interference. The self-interference noise calculator may include a network of processing elements configured to combine transmission signals into sets of intermediate results. Each set of intermediate results may be summed in the self-interference noise calculator to generate a corresponding adjusted signal. The adjusted signal is received by a corresponding wireless receiver to compensate for the self-interference noise generated by a wireless transmitter transmitting on the same frequency band as the wireless receiver is receiving.

Abstract: Systems, methods, and apparatuses related to cooperative learning neural networks are described. Cooperative learning neural networks may include neural networks which utilize sensor data received wirelessly from at least one other wireless communication device to train the neural network. For example, cooperative learning neural networks described herein may be used to develop weights which are associated with objects or conditions at one device and which may be transmitted to a second device, where they may be used to train the second device to react to such objects or conditions. The disclosed features may be used in various contexts, including machine-type communication, machine-to-machine communication, device-to-device communication, and the like. The disclosed techniques may be employed in a wireless (e.g., cellular) communication system, which may operate according to various standardized protocols.