| 研究生: |
廖偉廷 Wei-ting Liao |
|---|---|
| 論文名稱: |
使用動態階級化演算法之高階類神經電流模型 A High-Level Neural Current Model Using Dynamic Levelization Algorithm |
| 指導教授: |
劉建男
Chien-nan Liu |
| 口試委員: | |
| 學位類別: |
碩士 Master |
| 系所名稱: |
資訊電機學院 - 電機工程學系 Department of Electrical Engineering |
| 畢業學年度: | 97 |
| 語文別: | 中文 |
| 論文頁數: | 53 |
| 中文關鍵詞: | 動態階級化 、類神經 |
| 外文關鍵詞: | Dynamic Levelization, Neural |
| 相關次數: | 點閱:14 下載:0 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
現今VLSI製程技術已進入了奈米(Nanometer)的時代,使得電源完整性(Power Integrity)的問題成為限制電路效能原因之一。傳統上,要得到準確的電流波形須要到電晶體階層(Transistor level)做模擬才能得知,所以要到設計流程的後段才能去檢查電源完整性,因此,我們提出了使用動態階級化演算法(Dynamic levelization algorithm)之高階電流模型。動態階級化考慮了輸入向量以及對應的電流波形,把電路中的邏輯閘各分配到適當的Level來簡化電路複雜度,將各個Level的電流波形利用離散餘弦轉換(Discrete cosine transform)由時域波形轉為規律的頻域波形來處理。只需要一個模型,就可以估測不同形狀的電流波形,像是三角形波、梯形波亦或是多峰值波形,當模型建立好後,不須任何的手動調整,便可利用此模型來估測任意輸入向量所對應的電流波形。而實驗結果也證明,我們的高階電流模型確實可以產生足夠準確、相似度逼近電晶體階層的電流波形,提供使用者可以在設計初期做初步的雜訊估測。
As VLSI technology goes into nanometer era, the power integrity problem becomes one of the critical issues that limit the design performance. Traditionally, accurate supply current waveform can only be obtained from transistor-level simulation. Power integrity check is mostly performed at very late design stage in current design flow. Therefore, a high level current model for logic blocks using dynamic levelization algorithm is proposed to solve this problem. Proposed dynamic levelization algorithm can simplify the waveform complexity by grouping the gates by their logic level. Then, levelized current waveforms are transformed by DCT to obtain regular frequency domain waveforms. Thus, any shapes of current waveforms like triangular, trapezoidal, or multi-peak waveforms can be estimated by the proposed model without any manual tuning and extra information except the input/output values. Finally, the experimental results prove proposed that high-level current model provide accurate enough supply current waveforms for noise analysis at early design stages.
﹝1﹞Rishi Bhooshan, Bindu P Rao, “ Optimum IR Drop Models for Estimation of Metal Resource Requirements for Power Distribution Network ”, IFIP International Conference on Very Large Scale Integration, pp. 292-295, Oct. 2007
﹝2﹞Amir H. Ajami, Kaustav Banerjee, Amit Mehrotra, Massoud Pedram, “ Analysis of IR-Drop Scaling with Implications for Deep Submicron P/G Network Designs ”, International Symposium on Quality Electronic Design, pp. 35-40, Mar. 2003
﹝3﹞Yu Zhong, Martin D. F. Wong, “ Fast Algorithms for IR Drop Analysis in Large Power Grid ”, Computer-Aided Design, pp. 351-357, Nov. 2005
﹝4﹞” Simultaneous Switching Noise and Signal Integrity “, 2006 Actel Corporation
﹝5﹞Jin Zhao, Qing-lun Chen, ” A New Methodology for Simultaneous Switching Noise Simulation ”, Electrical Performance of Electronic Packaging, pp. 155-158, Oct. 2000
﹝6﹞Sang Won Song, Mohammed Ismail, Gyu Moon, Dong Yong Kim, ” Accurate Modeling of Simultaneous Switching Noise in Low Voltage digital VLSI ”, International Symposium on Circuits and Systems, Vol. 6, May 1999
﹝7﹞Atsushi Nakamura, Jyunichi Mano, Tatsuya Nagata, Hiroya Shimizu,Masayoshi Yagyu, Kunihiko Nishi, Kanji Otsuka,” Handling of Mutual Inductance in Simulation of Simultaneous Switching Noise ”, Electronic Components and Technology Conference, pp.663-668, May 1994
﹝8﹞Anand Raghunathan, Iraj K.Jha, Sujit Dey, High-LEVEL POWER ANALYSIS AND OPTIMIZATION, 1998
﹝9﹞Srinivas Bodapati, and Farid N. Najm,” High-Level Current Macro Model for Logic Blocks ” , IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Vol. 25, No. 5, May 2006
﹝10﹞楊侑成, “High-Level Current Macro Model using Neural Network”, June 2006
﹝11﹞羅華強,類神經網路-MATLAB的應用,六版,清蔚科技,民國九十年