Elimination of bottlenecks in dynamic dataflow processors

A key component of a dynamic dataflow processor, the matching unit, has been identified as a major bottleneck. An alternative implementation for the matching unit is presented. This implementation increases the operating bandwidth of the unit by allowing token-matching operations to be done concurrently. The adoption of the proposed implementation also reveals throughput limitations of other units within the processor. As a result, a configuration for a dynamic dataflow processor is proposed that is believed to be cost-effective, to give high performance, and to be capable of incremental performance improvement.<
>