Abstract: A data processing device provides for registers which can be formatted as segments containing numbers to which operations can be applied in SIMD fashion. In addition it is possible to perform operations which combine different segments of one register or segments at different positions in the different registers. By providing specially selected it is thus made possible to perform multidimensional separable transformations (like the 2-dimensional IDCT) without transposing the numbers in the registers.

Abstract: The disclosed data processing system comprises a memory means (SDRAM), a plurality of data processing means (IP) provided for accessing to said memory means (SDRAM), and a communication interface means coupled between said memory means (SDRAM) and said plurality of data processing means (IP), said communication interface means including a network of nodes (H 11, H 12, H2), each node comprising at least one slave port (s) for receiving a memory access request from a data processing means (IP) or from a previous node and at least one master port (m) for issuing a memory access request to a next node or to said memory means (SDRAM) in accordance with the memory access request received at said slave port (s), wherein said at least one slave port (s) is connected to a master port (m) of a previous node or to one of said data processing means (IP) and said at least one master port (m) is connected to a slave port (s) of a next node or to said memory means (SDRAM).

Abstract: Aspects involve effectively separating communication hardware in a data processing system by introducing a communication device for each processor. By introducing this separation the processors can concentrate on performing their function-specific tasks, while the communication device provide the communication support for the respective processor. Accordingly, in certain embodiments, a data processing system is provided with a computation, a communication support and a communication network layer.

Abstract: The dismissing of cached data that is not expected to be further used is predicted instead of predicting future I/O operations and then data is fetched from the main memory to replace the dismissed data in the cache. Thus, firstly a location in a cache memory containing data, which is expected not to be further used, is identified, followed by performing a prefetch operation in order to request new data to refill the above location in the cache memory. Therefore, a data processing system comprises at least one processor (12) for processing streaming data, at least one cache memory (200) having a plurality of cache blocks (210), wherein one of said cache memories (200) is associated to each of said processors (12), and at least one cache controller (300) for prefetching data into said cache memory (200), wherein one of said cache controllers (300) is associated to each of said cache memories (200).

Abstract: A data processing system according to the invention comprises a processor (P) and a memory hierarchy. The highest ranked level therein is a cache coupled to the processor. The memory hierarchy comprises a higher ranked cache (C1) having a cache controller (CC1) operating according to a write allocate scheme, and a lower ranked cache (C2) is coupled to the higher ranked cache (C1) having a cache controller (CC2). The size of the higher ranked cache is smaller than the size of the lower ranked cache. Both caches (C1, C2) administrate auxiliary information (V1, V2) indicating whether data (D1, D2) present therein is valid. The line size of the lower ranked cache (C2) is an integer multiple of the line size of the higher ranked cache (C1). The auxiliary information (V1) in the higher ranked cache (C1) concerns data elements (D1) at a finer granularity than that in the lower ranked cache (C2).

Abstract: A multiprocessor data processing system is described wherein the processors communicate to each other via a shared memory. Each of the processors comprises an administration unit (18a) and a computational unit. The administration unit of a writing processor maintains information defining a section in the memory which is free for storing data objects for readout by the reading processor. The administration unit of the reading processor maintains information defining a section in the memory in which the writing processor has written completed data for the data objects. The processors are arranged to signal a message to another processor via a processor synchronization channel for updating the information in the administration unit of said other processor.

Abstract: A data processing system is claimed which comprises a plurality of processors (12a, 12b, 12c) which communicate data streams with each other via a shared memory (10). The data processing system comprises processor synchronization means (18), for synchronizing the processors (12a-c) when passing the stream of data objects. For that purpose the processors are capable of issuing synchronization commands (Ca-c) to the synchronization means (18). At least one of the processors (12a) comprises a cache memory (184a), and the synchronization means (18) initiate a cache operation (CCa) in response to a synchronization commands (Ca).

Abstract: A data processing device provides for registers which can be formatted as segments containing numbers to which operations can be applied in SIMD fashion. In addition it is possible to perform operations which combine different segments of one register or segments at different positions in the different registers. By providing specially selected it is thus made possible to perform multidimensional separable transformations (like the 2-dimensional IDCT) without transposing the numbers in the registers.