摘要:
The power consumption of interconnects starts to have a significant impact on a system's total power consumption. Besides increasing buses (length, width) etc. this is mostly due to deep sub-micron effects where coupling capacitances between bus lines (wire-to-wire) are in the same order of magnitude as the base capacitances (wire-to-metal-layer). At that point, encoding schemes that solely address the minimization of transitions for the purpose of power reduction do not effectively work any more. Using a physical bus model that accurately models coupling capacitances, a signal bus encoding/decoding apparatus with encoding schemes that are partially adaptive and that take coupling effects into consideration is presented. The encoding schemes do not assume any a priori knowledge that is particular to a specific application.
摘要:
The apparatus comprises a discrete wavelet transform (DWT) engine, a code block manager, and an entropy encoder. The code block manager comprises at least one controller, which losslessly compresses the transform coefficients and stores them in a code block storage for buffering. The entropy coder comprises at least one entropy encoder, each comprising a decoder for decoding the losslessly compressed transformed coefficients prior to entropy encoding.
摘要:
According to one embodiment, a method for programming a programmable logic device (PLD) may include reading configuration data from a memory device to program a first portion of a PLD to function as a data decompression circuit (304, 308). Compressed configuration data may then be read and decompressed by the first portion and used to program a second portion (310, 312, 315) with a user determined function. A first portion may then be reprogrammed with a user determined function (320, 324).
摘要:
An apparatus and method for compressing pseudo-random data is provided. The apparatus and method make use of stochastic distribution models to generate approximations of the input data. A data sequence obtained from the stochastic distribution models is compared to the input data sequence to generate a difference data sequence. The difference data sequence tends to be less “random” than the input data sequence and is thus, a candidate for compression using pattern repetition. The difference data sequence is compressed using standard compression techniques and stored as a compressed data file along with information identifying the stochastic distribution model used and any parameters of the stochastic distribution model, including seed value and the like. When decompressing a data file compressed in the manner described above, the compressed difference data sequence is decompressed and a data sequence is generated using the identified stochastic distribution model and model parameters. The data sequence generated is then added to the difference data sequence to generate the original input data sequence.
摘要:
A coherent sampling method and apparatus in which synchronizing signal is synchronously obtained from an input signal. The synchronizing signal repetition time period is measured and thereby the clock cycle time, which executes a coherent sampling for the input signal, is sought. A clock generator generates the clock sought and because the sampling clock being optimized for the input signal is used, no waveform missing phenomena happens and a reproduced waveform is immediately displayable.
摘要:
A method for compressing an input sequence of data portions to produce an output codestream and for partially decompressing the output codestream to obtain a selected segment of the input sequence is disclosed. The output codestream includes a sequence of non-matchable sequences and codewords. The codewords include a first codeword and subsequent codewords. Each of the codewords includes at least a length of a subsequent non-matchable sequence preceding a matchable first sequence. Each of the subsequent codewords further includes a first offset for indicating a start of the matchable first sequence in the preceding non-matchable sequence, a length of the matchable first sequence and a second offset for indicating a location of a preceding codeword in the output codestream. A program storage device and a compressing/decompressing system for providing the above method are also disclosed.
摘要:
A data compression apparatus includes a source detection arrangement for detecting whether or not the input data is source data that has not undergone a previous compression/decompression cycle; a data quantity generator, responsive to the source detection arrangement, for setting a desired data output quantity for the compressed data, the desired data quantity having a first value for source input data and a second, higher, value for non-source input data; a target allocator for allocating a target data quantity to respective subsets of the input data in dependence upon the desired output data quantity, the target data quantities together providing a desired output data quantity; and a data compression arrangement for compressing each subset of the input data in accordance with its respective target data quantity.
摘要:
A matrix of AND-gates correspond to respective codes to be assigned to strings. The outputs of the AND-gates are selectively coupled to prefix code inputs of the AND-gates through a prefix code matrix switch. A plurality of characters from the input stream are fetched into an input character buffer and applied through respective character decoders and selectively through a character matrix switch to character inputs of the AND-gates. An AND-gate corresponding to a code assigned to a string that is the longest match to the plurality of fetched characters is thereby enabled. An update extended string is recorded in the AND-gate corresponding to the next code to be assigned by coupling, through the prefix code matrix switch, the output of the AND-gate corresponding to the longest matching string to the prefix code input of the AND-gate corresponding to the next code to be assigned. The character decoder output representing the fetched character following the longest matching string is coupled, through the character matrix switch, to the character input of the AND-gate corresponding to the next code to be assigned. A virtual level is assigned to an AND-gate indicative of the number of characters of the string recorded by the AND-gate. The virtual level that is assigned to the AND-gate recording the update extended string is one greater than the level that had been assigned to the AND-gate corresponding to the longest match. The virtual level is also indicative of the stage of the input character buffer holding the character following the longest match.
摘要:
An apparatus and method of calibrating a signal sourcing device having a plurality of signal sources to reduce errors between signals generated by the signal sources by categorizing the signal sources into groups based on the amount of error in the signal generated by the signal sources and selecting the signal sources from the groups in a manner which reduces the overall error in selecting signals. For example, the signal sourcing device may be a digital to analog converter having a plurality of current sources that output currents which may have an error. Such an improved digital to analog converter would categorize the current sources into groups based on the amount of error in the current generated by the current sources and selecting the current sources from the groups in a manner which reduces the overall error in selecting signals for the conversion process. The groups may include a center group that straddles the median current generated by the current sources. For example, the digital to analog converter may select current sources from complementary groups where current sources in complementary groups tend to have similar but opposite errors. This calibration method may be used to reduce errors in the signal sources which have a linear, Gaussian, or other error distributions.
摘要:
A method for data compression and decompression using a decompression language that consists of decompression instructions that are specified by using 3 bits to define which instruction and its syntax. There are nine decompression instructions, and each has a specific syntax, such as which character, or address length, position or distance, number of positions, etc. Two of the decompression instructions are specified by the same 3-bits and must be differentiated by a fourth bit. The compression method uses a character searcher that determines the locations for a particular character and produces optimum decompression instructions based upon repeated search results. The process is repeated for each unique character within the data. The compressed output may also be compressed by repeating the compression method until further compression can not be achieved or to maximum compression level 31. Decompression is accomplished by executing the decompression instructions until compression level zero is reached to reproduce the original data without any data loss.