30 results about "Vector size" patented technology

One embodiment of the present invention provides a processor that is configured to execute load-swapped-partial instructions. An instruction fetch unit within the processor is configured to fetch the load-swapped-partial instruction to be executed. Note that the load-swapped-partial instruction specifies a source address in a memory, which is possibly an unaligned address. Furthermore, an execution unit within the processor is configured to execute the load-swapped-partial instruction. This involves loading a partial-vector-sized datum from a naturally-aligned memory region encompassing the source address. While loading the partial-vector-sized datum, bytes of the partial-vector-sized datum are rotated to cause the byte at the specified source address to reside at the least-significant byte position within the partial-vector-sized datum for a little-endian memory transaction, or to cause the byte to be positioned at the most-significant byte position within the partial-vector-sized datum for a big-endian memory transaction.

A system and method Fast Fourier Transform (FFT) method in a multi-mode wireless processing system. The method can include loading an input vector into an input buffer, initializing a second counter and a variable N, where N=log2 (input vector size), and s is the value of the second counter, performing an FFT stage, and comparing s to N and performing additional FFT stages until s=N. Performing the FFT stage can include performing vector operations on data in the input buffer and sending results to an output buffer, the data in the input buffer comprising a plurality of segments, advancing the value of the second counter; and switching roles of the input and output buffers. The vector operations can include performing Radix-4 FFT vector operations on the four input data at a time and multiplying the resulting output vectors with a Twiddle factor.

Processing circuitry performs a plurality of lanes of processing on respective data elements of at least one operand vector to generate corresponding result data elements of a result vector. The processing circuitry identifies lane position information for each lane of processing, the lane position information for a given lane identifying a relative position of the corresponding result data element to be generated by the given lane within a corresponding result data value spanning one or more result data elements of the result vector. The processing circuitry is configured to perform each lane of processing in dependence on the lane position information identified for that lane. This enables generation of results which are wider or narrower than the vector size supported in hardware.

Provided are a QR decomposition apparatus and method for a MIMO system. The QR decomposition apparatus includes: a norm calculator for calculating a vector size norm for a channel input; a Q column calculator for calculating a column value of a unitary matrix Q by multiplying a delayed channel input with √{square root over (norm)}; an R row calculator for receiving the delayed channel input, the output of the Q column calculator, and 1 / √{square root over (norm)}, and calculating a row value of an upper triangular matrix R; a Q update calculator for receiving the delayed channel input, the output of the R row calculator, and a delayed output of the Q column calculator, and calculating a Q update matrix value; and a norm update calculator for receiving a delayed output of the norm calculator and an output of the R row calculator, and outputting a norm update matrix value.

This invention utilizes the moving vectors of last two correct mean absolute difference (MAD) matrixes to be the base of direction forecast and takes the last correct rector position as the center to forecast and fins out the possible direction of multiple MAD (min) phases and dimensions of vectors in the judgment way of extending layers by layer. Each time before comparing the next frame, a candidate block replaces the reference frame at the center, so that the shift vector varies in a certain sphere.

A binocular positioning method based on HOG feature and FAST algorithm, load the trained SVM model file; obtain the i-th frame image to be positioned, and copy the image midImage; preprocess the image midImage; use the FAST algorithm to perform spot Detect, obtain the coordinate vector points of the center point of the candidate area; judge whether the candidate area is the eye area in turn, and obtain the center point vector pointsTru of the eye area; filter the points in the vector pointsTru, and obtain the eye area whose final return vector size is at most 2 Center point vector; if only one eye is detected, repair it; return the center point vector pointsFnl of the eye area, traverse the vector pointsFnl, and intercept the rectangular area centered on pointsFnl as the eye area. The invention stably adapts to changes in the environment, has high robustness and accuracy to the environment, and has low false alarm rate.

The present invention relates to a method and apparatus for implementing a discrete Fourier transformation (DFT) of a predetermined vector size, wherein at least one enhanced DFT module is provided by using at least one type of DFT module including multiplication by first and second types of twiddle factors in respective different multiplication stages separated by an intermediate integration stage, and generating the enhanced DFT module by combining the at least one type of DFT module with a recursive stage configured to multiply by a third type of twiddle factor and to selectively switch between a bypass function and a butterfly function in said recursive stage. Thereby, an implementation of non 2x-radix Fourier transformation can be achieved with moderate hardware complexity.

The invention provides a single factor atmospheric polarization modeling method based on Rayleigh scattering, comprising the following steps: selecting the observation point and establishing a celestial coordinate system and obtaining the current time sun position using the observation time and observation point geographic coordinate; selecting any point to be measured on the celestial sphere surface and solely determining a northeast coordinate system based on the current time sun position using the point to be measured as the coordinate origin; in the northeast coordinate system, decomposing the incidence natural light into two beams of linearly polarized light in same intensity, vertical electric-field vibration direction, no fixed phase relation, and computing the scattered light electric field strength component under linearly polarized light incidence. The method uses the electric intensity vector to describe light wave and uses the vector size of the scattered light electric field strength component to obtain the polarization degree of the point to be measured and uses the vector direction of the scattered light electric field strength component to obtain the polarization direction of the point to be measured, thus the distributed model of the all-day atmospheric polarization information is obtained and the accuracy and the expansibility of the atmospheric polarization model is increased.

The invention discloses a river basin runoff forecast correction method and system. The method comprises the steps of obtaining real-time river basin runoff forecast data, corresponding rainfall data and runoff measured data, extracting historical data, calculating a correlation coefficient vector Vr of each set of historic runoff process data and other sets of historic runoff process data, dividing the historic runoff process data into two categories according to vectors, building a nonlinear forecast model for each category of the runoff process data, performing correlation analysis on the real-time river basin runoff forecast data, the corresponding rainfall data and the historic runoff process data, selecting the forecast model with the maximum coefficient as a correction model, adopting the real-time river basin runoff forecast data and the corresponding rainfall data as the input of the selected correction model, and outputting the corrected river basin runoff forecast data calculated by the selected correction model. According to the river basin runoff forecast correction method and system, the runoff forecast correction precision is improved, and the important decision information guidance is provided for reservoir optimal dispatching.

The invention discloses an FPGA-based SOPC multi-axis linkage motion controller and a control system. Among them, an FPGA-based SOPC multi-axis linkage motion controller includes a storage module, a soft-core IP and a multi-axis linkage motion control module. The multi-axis linkage motion control module is configured to execute the following program: set the continuous processing The starting processing point of the line segment is the origin under the X, Y, Z coordinate system; obtain the coordinate information of the current processing point in the line segment that can be processed continuously relative to the origin; obtain the virtual long axis of the vector in real time, among which, The vector direction of the virtual long axis is from the origin to the current processing point, and the vector size of the virtual long axis is the distance between the current processing point and the origin; the speed is adjusted in the vector direction of the virtual long axis, and then the virtual long axis is used to adjust X, Digital integral interpolation is performed on the Y and Z axes to realize the distribution of the feed pulses of the X, Y, and Z axes, and finally realize the speed regulation of the X, Y, and Z axes.

System and method embodiments are provided for adaptive vector size selection for vectorized query execution. The adaptive vector size selection is implemented in two stages. In a query planning stage, a suitable vector size is estimated for a query by a query planner. The planning stage includes analyzing a query plan tree, segmenting the tree into different segments, and assigning to the query execution plan an initial vector size to each segment. In a subsequent query execution stage, an execution engine monitors hardware performance indicators, and adjusts the vector size according to the monitored hardware performance indicators. Adjusting the vector size includes trying different vector sizes and observing related processor counters to increase or decrease the vector size, wherein the vector size is increased to improve hardware performance according to the processor counters, and wherein the vector size is decreased when the processor counters indicate a decrease in hardware performance.

The invention relates to a multi-level topic vector space construction method and device, an apparatus and a storage medium. The construction method comprises the following steps of extracting a global word co-occurrence matrix from a corpus; modeling the global word co-occurrence matrix, generating topic libraries of different levels, and respectively generating the topic correlation matrixes ofdifferent levels according to the topic libraries of different levels; and constructing a topic vector space for the topic correlation matrix of each level. Therefore, a topic can be expressed as thetopic vector of any dimension, and the defect that the vector size of the algorithms, such as LDA, etc., can only be the dictionary dimension, is avoided; meanwhile, based on the modeling of the noisewords, the influence of the noise words on clustering can be weakened.

Provided is a method and apparatus for receiving a signal for a MIMO system. The receiving apparatus includes: a QR decomposer for calculating a unitary matrix Q, an upper triangle matrix R, and a vector size for a received signal; a multiple dimension detector for calculating a first LLR for an output of the QR decomposer through multiple dimension detection; an inverse matrix and weight calculator for calculating an inverse matrix for the upper triangle matrix R and a weight; an interference remover for regenerating a symbol for a demodulated data stream using the fist LLR and removing interference from an output vector of the QR decomposer using the regenerated symbol; and a weight zero forcing unit for performing zero forcing on the interference removed output vector from the interference remover using the inverse matrix of the upper triangle matrix R and the weight and calculating a second LLR.