Abstract: A typical English pronunciation dictionary takes up to 1,826,302 bytes in ASCII to store. A five times compression while maintaining computability is achieved by prefix delta encoding of the word and error encoding of the pronunciation.

Abstract: A plug-and-play architecture including an extension of a general purpose processor media framework for adjoining DSP processing power.

Abstract: A class AD audio amplifier system (10) with DC output detection logic (26) is disclosed. The amplifier system (10) includes multiple audio channels (20), each of which includes a pulse-width-modulator (PWM) (24). The DC detection logic (26) includes a sigma-delta modulator (60) and a digital low-pass filter (62) that monitors the PWM output signals from the PWM modulators (24). The sigma-delta modulator (60) operates at a first clock frequency, while the low-pass filter (62) operates at a much lower clock frequency, so that AC audio components, PWM harmonics, and sigma-delta quantization error is suppressed from the DC detection. The modulated filtered signal is compared against a threshold level (THRSH) to determine whether the amplitude of a DC component at the PWM output is sufficiently high to constitute a fault. If so, a fault detection signal (DC_DET) is issued, and the PWM modulators (24) are disabled to prevent unsafe conditions in the system (10).

Abstract: An embodiment of the present invention includes a method of providing a non-uniform distribution of decoupling capacitors to provide a more uniform noise level across the chip. Leads on a packaged semiconductor chip are grouped into two or more regions. Types of leads needing decoupling capacitors are grouped into lead categories. For each region, there may be one or more lead categories therein. One or more decoupling capacitors are preferably assigned to each lead category in each region. Calculations may be performed to estimate a desired capacitance for each decoupling capacitor for each lead category in each region. When a chip has different components operating at different switching frequencies, different voltages, and/or different switching currents, the distribution of the decoupling capacitors will likely be non-uniform to provide a more uniform noise level across the chip, as compared to a uniform distribution of decoupling capacitors for the chip.

Abstract: A first periodic voltage waveform (20) is downconverted into a second periodic voltage waveform (35, 36). A plurality of temporally distinct samples (SA1, SA2, . . . ) respectively indicative of areas under corresponding half-cycles of the first voltage waveform are obtained. The samples are combined to produce the second voltage waveform, and are also manipulated to implement a filtering operation such that the second voltage waveform represents a downconverted, filtered version of the first voltage waveform.

Abstract: A micromirror capable of analog pulse width modulation, and method thereof. A capacitor (406) in each micromirror element stores a charge representative of one pixel of image data. A comparator (414) compares the image data signal stored on the capacitor (406) to a reference signal from a reference signal generator (416). Depending on the results of the comparison, the address electrodes (418) of the micromirror are biased to deflect the micromirror (412). As either the image data charge stored on the capacitor (406), or the reference voltage changes, the output of the comparator (414) changes to alter the position of the micromirror (412).

Abstract: The invention provides systems, devices, and methods for using more than one Accelerated Graphics Port (AGP) to process graphics for a single computer monitor (282). In one embodiment, the invention is a method of providing advanced/high-performance AGP capabilities to a laptop computer by intercepting AGP signals from a laptop having a low-power AGP (224) and converting the AGP signals to signals for a high-performance AGP (270).

Abstract: According to one embodiment of the invention, a method of fabricating ball grid array packages includes providing a substrate, providing a ball film that includes a plurality of metal balls movably contained within respective slots of a thin film, coupling the metal balls to the substrate, and removing the thin film from the metal balls.

Abstract: The present invention provides a method for placing a dopant in a substrate and a method for manufacturing an integrated circuit. The method for placing a dopant in a substrate, among other steps, includes providing a substrate (340) and implanting a dopant within the substrate (340) using an implant (370), the implant (370) moving at varying speeds across the substrate (340) to provide different concentrations of the dopant within the substrate (340).

Abstract: A method comprises performing at least one zero-bit scan across an interface. The at least one zero-bit scan defines a command window. Further, the method implements one of a selectable plurality of control levels in the command window based on the number of the at least one zero-bit scans.

Abstract: A method implemented in a test system comprises a test debug system and a target system, said target system comprising a test access port that functions according to a plurality of states and also comprising an adapter. The method comprises the adapter transferring data to the test debug system while the test access port remains in a predefined state. The predefined state comprises a state in which no scans occur.

Abstract: An identification (ID) process comprises in each of a plurality of bit times, a debug test system asserting a control signal at a predefined state to a plurality of target systems, and each target system, having a bit pattern and the bit patterns being different among the target systems, outputting a bit from its bit pattern on the control signal. The process further comprises each target system comparing the resulting state of the control signal to that target system's output bit. If the target system's output bit differs from the resulting control signal state, the target system ceases participating in the ID process or, if the target system's output bit matches the resulting control signal state, the target system continues to participate in the ID process.

Abstract: A method comprises performing at least one zero-bit scan across an interface link. The at least one zero-bit scan defines a command window. The method further comprises an interface adapter counting a number of inert scans in the command window, and the number of inert scans defines a particular command or data. An inert scan results in no data being moved into or out of the interface adapter.

Abstract: The present invention provides a method for manufacturing a semiconductor device and a method for manufacturing an integrated circuit. The method for manufacturing the semiconductor device, among other steps, includes forming a capping layer (210) over a transistor device having source/drain regions (150, 155) located over a substrate (110), the capping layer (210) having a degree of reflectivity, and annealing the transistor device through the capping layer (210) using photons (310), the annealing of the transistor device affected by the degree of reflectivity.

Abstract: A quantization method and system (20) for quantizing image data. A look up table (23) is used to provide non binary values for one or more bits of an input pixel value, as well as to provide a quantization error value. The quantization may be conditional, such that the look up table (23) is used to determine whether the input pixel value is to be quantized by the look up table (23) in non binary manner or whether it is to be quantized by a conventional binary quantizer (26). Where the look up table (23) quantizes only a portion of the input value, the remaining portion is delivered to the binary quantizer (26). The result is an output value, at least some of whose bits represent non binary values.

Abstract: A method is disclosed wherewith a person skilled in the art of statistical quality control may use to determine whether a process or a product is statically equivalent to another of known quality, or conforms to a desired known quality. The method may also be used to determine whether multiplicities of products or processes are statistically equivalent to one another. The method makes the determination based calculating an equivalency index that is associated with the measured parametric data, the population ratio of data points within a set of parametric limits.

Abstract: An emulation parameter indicative of a data processing operation performed by a data processor is exported from the data processor. The parameter value is provided as a plurality of digital bits. After determining that the bits of a first group within the plurality of bits all have the same bit value and that a predetermined bit within a second group of the plurality of bits has a bit value equal to the bit value of the bits of the first group, only the second group of bits is output from the data processor without outputting the first group of bits.

Abstract: A first data processor (GPP) can manage resources of a second data processor (DSP) by making a remote procedure call (RPC) to the second data processor to invoke on the second data processor a program that supports management of data processing resources of the second data processor. The second data processor executes the program in response to the remote procedure call.

Abstract: A semiconductor package comprising a die adjacent a lead frame die pad, said lead frame die pad adapted to dissipate heat from the die. The package further comprises a thermally-conductive material abutting the die and a heatsink abutting the thermally-conductive material, said heatsink facing a direction opposite from the lead frame die pad.

Abstract: A circuit (48) and method, which can be used in a mass data storage device, controls adaptation asymmetry of coefficients of an FIR filter (20) using an accumulator (52) or accumulating correlation results between unequalized FIR filter input data samples and FIR filter output equalized error samples. A circuit (52) generates coefficient increment and decrement requests from the accumulated correlation results. A circuit (120,102?,122) updates the coefficients within a symmetric coefficient pair on the basis of the increment and decrement requests only if a predetermined nonzero coefficient magnitude difference between the coefficient pair would not be exceeded by the update.