A scheduling algorithm for conditional resource sharing-a hierarchical reduction approach

Using speculative computation and parallelizing techniques to improve scheduling of control based designs

“…A first subset is composed of small benchmarks derived from [19] (sehwa), [20] (maha) and [11] (kim), while the second one is composed of a set of multimedia applications extracted from the Mediabench suite [13] The results for each technique and each target benchmark are shown in terms of the computational time needed to generate the schedule and the number of control steps. The computational times are average values obtained by applying the scheduling processes for ten times.…”

Coordinated parallelizing compiler optimizations and high-level synthesis

“…Early high-level synthesis work focused on data-flow designs and applied optimizations such as algebraic transformations, re-timing and code motions across multiplexors for improved synthesis results [13,14]. More recent work -during the last decadehas presented speculative code motions for mixed control-data flow type of designs and demonstrated their effects on schedule lengths [15,16,17,18,19,20,21,22,23].…”

Hierarchical conditional dependency graphs as a unifying design representation in the CODESIS high-level synthesis system

“…Finally, conditional resource sharing is exploited during scheduling and not before and so lengthening of execution paths due to inappropriate conditional resource sharing (i.e. [2], [11]), is also avoided.…”

“…These negative effects intervene in two distinct but interrelated levels as far as scheduling conditional behaviors is concerned; mutual exclusiveness detection and operation scheduling. CDFG based mutual exclusiveness detection techniques [3], [11] using the structure of the input description, produce different schedules for semantically equivalent but syntactically different descriptions. This is due to the variability on the amount of detected mutual exclusiveness [9].…”