在此論文當中,針對邏輯匣尺寸和閥值電壓的選派任務,提出一個漏電流最小化的設計方法。在現今電子產品強調高效能和低耗能的趨勢下,節省電能的消耗已是熱門議題,其中漏電流的節約更是重要的因素。因此選用適當的邏輯匣尺寸和閥值電壓,不僅能夠達到高效能的需求,同時可以降低漏電流值。 為了達到漏電值的最小化,本論文以模擬退火法為基礎的演算法,結合遞迴性演算法作為初始值。另外,為探索足夠的解空間及尋找最佳化的解,實作計時器以快速計算訊號到達時間與需求時間的差異值。實驗結果顯示,此設計方法能夠在效能限制的條件中有效降低漏電流值。
We propose an iterative (IR) and simulated-annealing (SA) based methodology for leakage power minimization by the means of gate sizing and threshold voltage assignment. Currently, there is a strong market for high-performance mobile devices to have low power consumption. Among many low power consumption oriented design techniques, leakage power minimization is one of the key technologies. For cell-based design, gate sizing and threshold voltage assignment can be employed to meet timing requirement while reducing total leakage power. We apply simulated-annealing-based optimization on an initial solution that is generated by an iterative algorithm. In order to explore more solution space, we implement a timing engine for fast calculation of timing slack. Experimental results show that the proposed methodology can effectively reduce leakage power under performance constraints.