Abstract: The present invention provides alignment and ordering of vector elements for SIMD processing. In the alignment of vector elements for SIMD processing, one vector is loaded from a memory unit into a first register and another vector is loaded from the memory unit into a second register. The first vector contains a first byte of an aligned vector to be generated. Then, a starting byte specifying the first byte of an aligned vector is determined. Next, a vector is extracted from the first register and the second register beginning from the first bit in the first byte of the first register continuing through the bits in the second register. Finally, the extracted vector is replicated into a third register such that the third register contains a plurality of elements aligned for SIMD processing. In the ordering of vector elements for SIMD processing, a first vector is loaded from a memory unit into a first register and a second vector is loaded from the memory unit into a second register.

Abstract: The present invention provides extended precision in SIMD arithmetic operations in a processor having a register file and an accumulator. A first set of data elements and a second set of data elements are loaded into a first vector register and a second vector register, respectively. Each data element comprises N bits. Next, an arithmetic instruction is fetched from memory. The arithmetic instruction is decoded. Then, a first vector register and a second vector register are read from the register file. The present invention then executes the arithmetic instruction on corresponding data elements in the first and second vector registers. The result of the execution is then written into the accumulator. Then, each element in the accumulator is transformed into an N-bit width element and stored into the memory.

Abstract: A mechanism and method for recording stereo video with standard camera system electronics and a uniquely adapted optical assembly is disclosed. The optical assembly comprises left and right optical channels disposed to capture and project separate left and right images onto a single image sensor such that the boundary between the projected images is sharply delineated with no substantial overlap or gap. The viewpoints of the left and right optical channels are separated by a distance, d, such that the captured images are differentiated to produce a stereo image pair. By proper disposition of the left and right optical channels, stereo image pairs exhibiting full stereo overlap without keystone distortion are obtained. One image of the stereo pair is produced for visualization by the left eye and the other image is produced for visualization by the right eye. Alternatively, the images can be interrogated by a computer system for generating three dimensional position data.

Abstract: A method and arrangement for separating interleaved luminance and chrominance color space components data in a single data stream with minimum CPU intervention is provided. In the separating circuit, the separating circuit receives as input a series of graphics/video image data composed of interleaved luminance and chrominance color space components at successive clock cycles. The separating circuit directs selected bytes of the graphics/video image data representing the luminance color space component to a first path wherein luminance component data received at two successive clock cycles are combined. Likewise, selected bytes of the graphics/video image data representing the chrominance color space component are directed to a second path wherein chrominance component data received at two successive clock cycles are combined. Then, the combined luminance and chrominance component data are output alternately.

Abstract: Instruction methods for moving data between memory and a vector register file while performing data formatting. The methods are processed by a processor having a vector register file and a memory unit. The methods are useful in the graphics art because they allow more efficient movement and processing of raster formatted graphics data. The vector register file has a number of vector registers (e.g., 32) that each contain multi-bits of storage (e.g., 128 bits). In one class of instructions, eight byte locations within memory are simultaneously loaded into eight separate 16 bit locations within a register of the register file. The data can be integer or fraction and signed or unsigned. The data can also be stored from the register file back to memory. In a second class of instructions, alternate locations of a memory qaudword are selected and simultaneously loaded in the register file.

Abstract: This relates to a general purpose circuit that maximizes the computing power of a Unix workstation or other computer system for processing image or other data in accordance with a selected one or ones of several alternative compression and decompression algorithms. This dynamically allocates system memory for storage of both compressed and uncompressed data and ensures adequate compression and decompression rates.

Abstract: Methods for forming computer models of curves, networks, or surfaces from user defined specifications of the shape to be modeled. Each specification includes a set of geometric constraints, such as positions, tangents curvatures, and torsions, and may also include discontinuity specifications. In the preferred embodiment, curves are computed so as to locally minimize a scale invariant functional of the geometry of the curve, such as a magnitude of variation in curvature of the curve (MVC) or a magnitude of curvature of the curve (MEC), while satisfying a user defined specification. An improvement on the MVC functional is to add a magnitude of variation in torsion of the curve. An improvement on the MEC functional is to add a magnitude of torsion of the curve.

Abstract: A computer-implemented method of transmitting images from a transmitter to a receiver (e.g. in a teleconferencing application). A receiver maintains an image in a local storage (e.g. that from a previous frame in a sequence of frames) and the transmitter receives an updated image for a next temporal period (e.g. the next frame). The transmitter divides the updated image into blocks and comparing a rotating pixel sample(s) of each of the blocks from the updated image with a sampled pixel from a local copy of a receiver's image at a same spatial position of the pixel sample(s). The transmitter determines a difference between the rotating sampled pixel of each of the blocks from the updated image and the local copy of the receiver's image. It stores a reference to the block and associates the difference with the reference.

Abstract: A computer-implemented method of transmitting images from a transmitter to a receiver (e.g. in a teleconferencing application). A receiver maintains an image in a local storage (e.g. that from a previous frame in a sequence of frames) and the transmitter receives an updated image for a next temporal period (e.g. the next frame). The transmitter divides the updated image into blocks and comparing a rotating pixel sample(s) of each of the blocks from the updated image with a sampled pixel from a local copy of a receiver's image at a same spatial position of the pixel sample(s). The transmitter determines a difference between the rotating sampled pixel of each of the blocks from the updated image and the local copy of the receiver's image. It stores a reference to the block and associates the difference with the reference. The difference is an average absolute difference in luminance between the two blocks.