Abstract: A system and method are provided that include determining optimum memory organization in an electronic device, wherein further determined are optimum resource interconnection patterns. One aspect of the system and method includes determining resource, e.g., memories and data paths, interconnection patterns of complex bus structures with switches using system-level information about the data-transfer conflicts. The quantity of memories within an electronic device, the size of the memories and the interconnection between the memories, including the interconnection of the memories with one or more data paths, defines a memory organization of an electronic device. Another aspect of the system and method relates to selecting an optimized memory organization, including selecting an optimized interconnection pattern between the memories and between the memories and the data paths.

Abstract: A system and method for determining optimized scheduling intervals and optimized access conflicts and for determining an optimized memory organization of an essentially digital device. The system includes an optimizer for determining an optimized scheduling of the data access instructions for a plurality of disjunct code blocks, wherein each of the code blocks include part of the data access instructions. The system performs an iterative process of successively reducing the cycle budget for selected blocks and modifying the scheduling of the selected blocks until a cumulative cycle budget for all of the blocks is met.

Abstract: A system and method of designing digital system. One aspect of the invention includes a method for designing an essentially digital system, wherein Pareto-based task concurrency optimization is performed. The method uses a system-level description of the functionality and timing of said digital system, wherein the system-level description comprises a plurality of tasks. Task concurrency optimization is performed on said system-level description, thereby obtaining a task concurrency optimized system-level description, including Pareto-like task optimization information. The essentially digital system is designed based on said task concurrency optimized system-level description. In one embodiment of the invention, the description is includes a “grey-box” description of the essentially digital system.

Abstract: One aspect of the invention includes a method of address expression optimization of source-level code. The source-level code describes the functionality of an application to be executed on a digital device. The method comprises first inputting first source-level code that describes the functionality of the application into optimization system. The optimization system then transforms the first source-level into a second source level that has fewer nonlinear operations than the first source-level code.

Abstract: A formalized method and a design system are described for part of the design decisions, related to memory, involved while designing an essentially digital device. The method and system determine an optimized memory organization starting from a representation of said digital device, the representation describing the functionality of the digital device and comprising data access instructions on basic groups, which are groups of scalar signals. The method and system determine optimized scheduling intervals of said data access instructions such that execution of said functionality with the digital device is guaranteed to be within a predetermined cycle budget, the determining of the optimized scheduling intervals comprising optimizing access conflicts with respect to an evaluation criterion related to the memory cost of said digital device. An optimized memory organization is selected in accordance with the optimized scheduling intervals and the optimized access conflicts.

Abstract: A system and method for determining optimized scheduling intervals and optimized access conflicts and for determining an optimized memory organization of an essentially digital device. The system includes an optimizer for determining an optimized scheduling of the data access instructions for a plurality of disjunct code blocks, wherein each of the code blocks include part of the data access instructions. The system performs an iterative process of successively reducing the cycle budget for selected blocks and modifying the scheduling of the selected blocks until a cumulative cycle budget for all of the blocks is met.

Abstract: A method for determining an optimized data organization in at least one first memory of an essentially digital system comprising at least the first memory and a second memory, acting as cache for the first memory, the optimized data organization being characteristic for an application, to be executed by the digital system, is presented.

Abstract: A programmable processing engine and a method of operating the same is described, the processing engine including a customized processor, a flexible processor and a data store commonly sharable between the two processors. The customized processor normally executes a sequence of a plurality of pre-customized routines, usually for which it has been optimized. To provide some flexibility for design changes and optimizations, a controller for monitoring the customized processor during execution of routines is provided to select one of a set of pre-customized processing interruption points and for switching context from the customized processor to the flexible processor at the interruption point. The customized processor can then be switched off and the flexible processor carries out a modified routine. By using sharable a data store, the context switch can be chosen at a time when all relevant data is in the sharable data store. This means that the flexible processor can pick up the modified processing cleanly.

Abstract: The present invention provides a method and an apparatus for reducing the storage size required for temporary data by storage order optimization. Advantageously, the execution order optimization and the storage order optimization may be treated independently. The storage size optimization is preferably performed by determining an optimum intra-array and/or inter-array storage order based on a geometrical model. The geometrical model provides a representation of the address space occupied by an array as a function of time and allows the calculation of the window size of the occupied address/time domain of the array. Where calculations would be time-consuming, these may be shortened by making simplifying assumptions, e.g. calculation of upper and lower bounds of the window size of the occupied address/time domain of an array rather than an exact calculation. Further, heuristical simplifications are described to reduce run-times for the optimization process.

Abstract: Selected code is modeled in a polyhedral dependency graph (PDG). A placement optimizer maps each element of the PDG to an optimally placed PDG. An ordering optimizer maps the placed PDG to an optimally ordered PDG. The PDG, place PDG, and ordered PDG are combined to produce a transformation script. The transformation script is applied to the selected specification description to produce optimized selected code. Optimized selected code is combined with original code to generate a control-flow optimized code. In addition, memory directives are derived from the ordered PDG model. The memory directives and optimized code are used to generate target code for simulation or software compilation.

Abstract: A battery-powered computing system (20) including video decoding capability, particularly as pertinent to the H.263 standard, is disclosed. The system (20) includes a main integrated circuit (30) having an on-chip central processing unit (CPU) (32) and on-chip shared memory (33) for the temporary buffering of video image data that is retrieved and generated during the video decoding process. The CPU (32) is programmed to perform a combined P and B prediction process (46) upon a previously predicted P frame (P.sub.T-1), with accesses to internal buffers in shared memory (33) instead of to main memory (40). Preferably, inverse transform processes (48, 52) also access shared memory (33) rather than main memory (40). The combined P and B prediction process (46) preferably handles unrestricted motion vectors using edge pixels (P.sub.

Abstract: Data storage and transfer cost is responsible for a large amount of the VLSI system realization cost in terms of area and power consumption for real-time multi-dimensional signal processing applications. Applications or this type are data-dominated because they handle a large amount of indexed data which are produced and consumed in the context of nested loops. This important application domain includes the majority of speech, video, image, and graphic processing (multi-media in general) and end-user telecom applications. The present invention relates to the automated allocation of the background memory units, necessary to store the large multi-dimensional signals. In order to handle both procedural and nonprocedural specification, the novel memory allocation methodology is based on an optimization process driven by data-flow analysis. This steering mechanism allows more exploration freedom than the more restricted scheduling-based investigation in the existent synthesis systems.