Abstract: An operating method and a circuit for low density parity check (LDPC) decoders, in which original bit nodes are incorporated into check nodes for simultaneous operation. The bit node messages are generated according to the difference between the newly generated check messages and the previous check node messages. The bit node messages can be updated immediately, and the decoder throughput can be improved. The required memory of LDPC decoders can be effectively reduced, and the decoding speed can also be enhanced.

Abstract: A multi-mode multi-parallelism data exchange method and the device thereof are proposed to apply to a check node operator or a bit node operator. The proposed method comprises the steps of: duplicating part or all of an original shift data as a duplicated shift data; combining the original shift data and the duplicated shift data to form a data block; and using a data block as the unit to shift this data block so as to conveniently retrieve shift data from the shifted data block. With a maximum z factor circuit and duplication of part of data, specifications of different shift sizes can be supported. The functions of shifters of several sizes can therefore be accomplished with the minimum complexity.

Abstract: A multi-mode multi-parallelism data exchange method and the device thereof are proposed to apply to a check node operator or a bit node operator. The proposed method comprises the steps of: duplicating part or all of an original shift data as a duplicated shift data; combining the original shift data and the duplicated shift data to form a data block; and using a data block as the unit to shift this data block so as to conveniently retrieve shift data from the shifted data block. With a maximum z factor circuit and duplication of part of data, specifications of different shift sizes can be supported. The functions of shifters of several sizes can therefore be accomplished with the minimum complexity.

Abstract: An operating method applied to low density parity check (LDPC) decoders and the circuit thereof are proposed, in which original bit nodes are incorporated into check nodes for simultaneous operation. The bit node messages are generated according to the different between the newly generated check messages and the previously check node messages. The bit node messages can be updated immediately, and the decoder throughput can be improved. In the other way, the required memory of LDPC decoders can be effectively reduced, and the decoding speed can also be enhanced.

Abstract: A method for assuring virtual atomic invalidation in a multilevel cache system wherein lower level cache locations store portions of a line stored at a higher level cache location. Upon receipt of an invalidation signal, the higher level cache location invalidates the line and places a HOLD bit on the invalidated line. Thereafter, the higher level cache sends invalidation signals to all lower level caches which store portions of the invalidated line. Each lower level cache invalidates its portion of the line and sets a HOLD bit on its portion of the line. The HOLD bits are reset after all line portion invalidations have been completed.

Abstract: A method of maintaining cache coherency in a shared memory multiprocessor system having a plurality of nodes, where each node itself is a shared memory multiprocessor. With this invention, an additional shared owner state is maintained so that if a cache at the highest level of cache memory in the system issues a read or write request to a cache line that misses the highest cache level of the system, then the owner of the cache line places the cache line on the bus interconnecting the highest level of cache memories.

Abstract: A method of reducing false sharing in a shared memory system by enabling two caches to modify the same line at the same time. More specifically, with this invention a lock associated with a segment of shared memory is acquired, where the segment will then be used exclusively by processor of the shared memory system that has acquired the lock. For each line of the segment, an invalidation request is sent to a number of caches of the system. When a cache receives the invalidation request, it invalidates each line of the segment that is in the cache. When each line of the segment is invalidated, an invalidation acknowledgement is sent to the global directory. For each line of the segment that has been updated or modified, the update data is written back to main memory. Then, an acquire signal is sent to the requesting processor which then has exclusive use of the segment.

Abstract: A cache coherence protocol for a multiprocessor system. Each processor in the system has an associated cache capable of storing multiple word data lines. The system also includes a plurality of main memory modules, each having an associated distributed global directory storing directory information for lines stored in the associated main memory module. Each main memory module is connected to each processor by means of a multi-stage interconnection network. When a processor attempts to over-write an individual word in a line stored in its associated cache, a write request signal is sent to the appropriate global directory, and each other processor whose cache stores a copy of the line is notified of the request. When each other processor has responded with an acknowledgement, the first processor is allowed to proceed with the write.

Abstract: An optimization scheme for a directory-based cache coherence protocol for multistage interconnection network-based multiprocessors improves system performance by reducing network latency. The optimization scheme is scalable, targeting multiprocessor systems having a moderate number of processors. The modification of shared data is the dominant contributor to performance degradation in these systems. The directory-based cache coherence scheme uses an invalidation bus on the processor side of the network. The invalidation bus connects all the private caches in the system and processes the invalidation requests thereby eliminating the need to send invalidations across the network.

Abstract: A switch queue structure for one-network parallel processor systems minimizes chip count and reduces the possibility of deadlock which might otherwise occur with this type of switch structure. The switch queue structure comprises a plurality of input ports and a plurality of output ports equal in number to a number of processor/memory elements (PMEs) in a parallel processor system. A plurality of identical stages interconnect the plurality of input ports and the plurality of output ports. Each stage includes a plurality of first groups of first-in, first-out (FIFO) registers storing request messages, a plurality of second groups of first-in, first-out registers storing response messages, and a plurality of multiplexers.

Abstract: A system and method for a fault-tolerant system for parallel networks which interconnect processors and the first of the parallel networks distributed in an Omega configuration and the second of the parallel networks distributed in a reversed Omega configuration.

Abstract: A combining switch 10 includes a two input multiplexer 12 which receives I and J inputs from data processors and directs one of the incoming messages, if there are no contentions of congestions at a switch output port 14 and a Quene FIFO 16 is empty, directly to the output port 14 for transmission to one of a plurality of memory modules. If the output port 14 is busy and the Queue 16 is empty the incoming message is routed to the Queue FIFO 16 for storage. If the Queue FIFO 16 is not empty the incoming message is first compared by a comparator 20 to all existing messages stored in the Queue FIFO 16 to determine if the incoming messasge is destined for a memory address which already has a queued message. If no match is determined by comparator 20 the incoming message is routed to the Queue FIFO 16 for storage.

Abstract: Using a variable-duration clock circuit, together with programmable duration control to alter the clock waveform within strict rules, permits the programmer to arrange appropriately short durations for short data transfers, and to arrange appropriately longer durations for longer data transfers in an array processor of myriad processing elements. There is no need to allow sufficient time in every clock cycle for worst case data transfer between remote processing elements.The clock waveform has three recognizable edges (A,B,C) regardless of loss of sharpness during its travel to the various processing elements. The convention that three skew-sensitive activities, READ, WRITE and OPERAND SUPPLY conform to respectively assigned edges as follows:A=READ;B=OPERAND SUPPLY;C=WRITE (Read next)The processing elements synchronize with the clock waveform, which is optimized for the instructions of the program being executed.