行路由PEA广度贪心调度映射算法

发布时间：2019-08-26 10:03

【摘要】：粗粒度可重构单元阵列硬件任务的贪心映射是可重构计算要解决的核心问题。不同的阵列具有不同的硬件约束条件,针对行路由粗粒度可重构单元阵列提出一种广度贪心映射算法BGMA(Breadth Greedy Mapping Algorithm)。该算法首先从第一个节点开始依次扫描,如果节点满足条件则将其映射到PEA上,当遇到不满足映射条件的节点时,该算法将跳过该节点继续寻找满足约束条件的节点进行映射,通过与广度不贪心映射算法BNGMA(Breadth No Greedy Mapping Algorithm)相比较,BGMA的N1平均减少了35.1%(PEA_(6×6))和54.8%(PEA_(8×8)),N2平均减少了35.6%(PEA_(6×6))和54.6%(PEA_(8×8)),C_(CON)平均减少了15.7%(PEA_(6×6))和26.2%(PEA_(8×8)),T_(TOTAL)平均减少了20.2%(PEA_(6×6))和32.1%(PEA_(8×8))。实验结果表明了贪心策略在映射算法中的重要性。
[Abstract]:Greedy mapping of hardware tasks of coarse-granularity reconfigurable cell array is the core problem to be solved in reconfigurable computing. Different arrays have different hardware constraints. A breadth greedy mapping algorithm BGMA (Breadth Greedy Mapping Algorithm). Is proposed for row routing coarse granularity reconfigurable cell arrays. The algorithm starts with the first node and maps it to PEA if the node satisfies the condition. When the node does not meet the mapping condition, the algorithm will skip the node and continue to find the node that satisfies the constraint condition to map, by comparing with the breadth non-greedy mapping algorithm BNGMA (Breadth No Greedy Mapping Algorithm). The N1 of BGMA decreased by 35.1% (PEA_ (6 脳 6) and 54.8% (PEA_ (8 脳 8), N2) by 35.6% (PEA_ (6 脳 6) and 54.6% (PEA_ (8 脳 8), C _ (CON) by 15.7% (PEA_ (6 脳 6) and 26.2% (PEA_ (8 脳 8), T _ (TOTAL) by 20.2% (PEA_ (6 脳 6) and 32.1% (PEA_ (8 脳 8). The experimental results show the importance of greedy strategy in mapping algorithm.
【作者单位】：安徽工程大学计算机与信息学院;
【基金】：安徽省自然科学基金(No.1408085MF124) 安徽省高校省级自然科学基金重点项目(No.kj2015A003) 安徽工程大学国家自然科学预研基金
【分类号】：TP301.6;TP332

【相似文献】