Hardware binding plays an important role in the performance of a design on FPGAs. Good timing performance requires that the hardware binding be as efficient as possible. It is often acceptable to let the area increase...
详细信息
ISBN:
(纸本)9780769544472
Hardware binding plays an important role in the performance of a design on FPGAs. Good timing performance requires that the hardware binding be as efficient as possible. It is often acceptable to let the area increase within a tolerance limit if the timing could be improved. In this paper, we propose a new hardware binding algorithm.. It performs simultaneous FU and register binding incorporating device-specific delay information for functional units and multiplexers. The proposed approach, implemented within a C to RTL framework has resulted in significant improvement in maximum achievable clock frequency compared to previously proposed weightedbipartitematching and Compatibility Path Based algorithms. The associated increase in area is also within a very tight margin and hence quite acceptable even when there is an area constraint. Also, when compared to WBM and CPB methods, the proposed algorithm improves clock period on average by 17.6% and 9.7% respectively without any penalty in area. When compared with ECPB algorithm, clock period is improved by 5.6% on average at a small area cost of 5.5%.
This paper addresses unrelated parallel machine scheduling problems with two minimization objectives: total weighted flow time and tardiness, and presents two hybrid methods based on (1) non-dominated sorting genetic ...
详细信息
This paper addresses unrelated parallel machine scheduling problems with two minimization objectives: total weighted flow time and tardiness, and presents two hybrid methods based on (1) non-dominated sorting genetic algorithms (NSGA-II) and (2) strength Pareto evolutionary algorithm (SPEA). These algorithms were implemented in a different manner according to the following two features: (1) using random or fixed weighted sum direction search (RWSD or FWSD); (2) including or not including a bipartiteweightedmatching problem (BWMP). The performance of the algorithms is evaluated via two benchmark instances generated by a method in the literature. The experimental results indicate that algorithms with RWSD are superior to those with FWSD, and those including BWMP outperforms those not, in terms of proximity and spread metrics. In particular, NSGA-II with RWSD and BWMP performs best for the large size instance, whereas SPEA with RWSD and BWMP excels other algorithms in solving the medium size instance. Nevertheless, algorithms without BWMP spend much less computation time than others under the same termination criterion
暂无评论