臺灣博碩士論文加值系統

摘要
低功率設計在現今的超大型積體電路(VLSI)中已成為舉足輕重的一個環節。由於匯流排通常消耗大量的功率，減少晶片功率消耗的有效方法是降低匯流排上的功率消耗。因此，匯流排編碼(bus coding)方法是利用匯流排資料特性來降低匯流排上資料轉換的次數(switching activity)以及在匯流排靜態電流的流失，因而達到低功率消耗的設計目的。 在積體電路中，主要的功率消耗包含有負載電容，活動開關，工作頻率以及供應電壓。每個工程師都必須依據上面幾點功率消耗的因素，解決在不同的設計階層關於功率消耗的問題其中包含有電路(circuit)層、架構(architecture)層、邏輯(gate)層等。基本上,功率消耗是與負載電容成比例關係。此外,外部匯流排(off-chip)上的負載電容約為內部匯流排(on-chip)的一千倍。相對地，每次開關活動(switching activity)的變化，在匯流排(off-chip)上的功率消耗約為內部匯流排(on-chip)的一千倍。假如有效地降低匯流排上的開關活動(switching activity)將能夠減少功率消耗。在這論文中，為了有效減少在匯流排上的開關活動，我們提出一種新的編碼技術。此外，它也可以節省動態功率的消耗。這編碼方法命名為XOR-BITS碼。這XOR-BITS碼可同時被運用在位址匯流排以及在資料匯流排上。它能在位址匯流排上減少74.16%的開關活動。同時，它也能在資料匯流排上減少13.65% 的開關活動。因此，這XOR-BITS碼不僅節省較多的功率消耗並且提升整體系統上的效能。 我們的編碼法與其他編碼法相互比較。XOR-BITS碼被證實比其他編碼法有較小的開關活動在資料匯流排以及位址匯流排上。此外, XOR-BITS碼的電路複雜度並不會與其他編碼法電路增加許多。我們使用MATLAB軟體去計算開關活動(switching activities)數。此外，我們也使用Modelsim 軟體驗證功能與Design Vision軟體去合成我們的設計，並且估計它的邏輯數與功率消耗。

In modern VLSI design, low power design is one of the key issues. Since on-chip and off-chip bus consume a general quantity of power, one of the effective methods to reduce the power consumption on chip is to reduce the power consumed by the on-chip and off-chip buses. Therefore, bus coding methods, that make capital of bus value characteristics to reduce the switching activity and static current loss on a bus, have been proposed to accomplish the objective. In integrated circuit, major sources for power dissipation include load capacitance, switching activity, work frequency and supply voltage. Based on the above power dissipation factors, every engineer has to solve power dissipation question in different design levels, including circuit level, architecture level, and gate level. The power dissipation, in the main, is in proportion to load capacitance. Moreover, the load capacitance from off-chip buses is about thousands of that from on-chip buses. Relatively, the power dissipation from off-chip buses is about thousands of that from on-chip buses. If we are effectively lowered the switching activity of the bus, we can reduce power dissipation significantly. In this thesis, the purpose of the proposed new coding technique is to effectively diminish the switching activity on the busses. In addition, it can save dynamic power consumption. The proposed encoding method is called the XOR-BITS code. The XOR-BITS coding can be used on address bus and it can also be applied on data bus simultaneously. It can reduce about 74% switching activity in the address bus. Meanwhile, it can also reduce switching activity in the data bus for about 14%. Consequently, the proposed XOR-BITS code can not only save more power consumption but also extend the whole system working time. Our bus coding method contrasts mutually with other bus coding method. The XOR-BITS coding is verified to achieve less switching activity than other coding methods on data bus and address bus. Besides, the increase of the circuit complexity is limited. We have used the MATLAB© software to calculate the switching activities in behavior simulation. Furthermore, we have used the Design Vision© software to synthesize our bus coder, and estimate the gate counts and power consumption.

Content
Chapter 1.Introduction ........................................1
1.1 Background ................................................1
1.2 Source of Power Consumption ...............................3
1.3 Designing Level of the Low Power ..........................6
1.4 Outline of This Paper .....................................7 Chapter 2.Bus Coding Method ...................................9
2.1 Basic Notion of the Low Power Bus .........................9
2.2 Characteristic Interpretation of the Busses ...............12 2.3 Coding Methods for Data Bus ...............................14 2.3.1 Bus-Invert (BI) Method ..................................14 2.3.2 Bus-invert transition signaling (BITS) Method ...........16 2.3.3 Half-Identity Half-Reverse Transition Signaling (HihrTS) Method .16
2.3.4 Partial Bus-Invert (PBI) Method .........................17 2.4 Coding Methods for Address Bus ............................18 2.4.1 Gray Method .............................................18 2.4.2 T0 Method ...............................................19 2.4.3 T0-Xor Method ...........................................21
2.4.4 ALBORZ Method ...........................................23 2.4.5 Irredundant ALBORZ Method ...............................24 2.5 The Hardware Comparison of Different Coding Methods .......25 Chapter 3.Novel Mixed Bus Encoding Technique ..................28 3.1 Consideration of Designed Principle .......................28 3.2 The New Encoding Algorithm ................................30 3.3 The New Decoding Algorithm ................................38 Chapter 4.Simulation, Verification and Performance Comparisons 45 4.1 Environment of Simulation .................................45 4.2 Method of Estimation ......................................46 4.3 Simulation Result .........................................48 4.3.1 Various Kinds of Simulation Methods .....................48 4.3.2 Estimation and Contrast .................................52 Chapter 5.Conclusions and Future Work .........................64 5.1 Conclusions ...............................................64 5.2 Future Work ...............................................66 Reference .....................................................67

6. References
[1] M.R. Stan and W.P. Burleson, “Bus-invert coding for low-power I/O,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Volume: 3 , Issue: 1 , pages:49 – 58, March 1995.
[2] Youngsoo Shin and Kiyoung Choi and Young-Hoon Chang,” Narrow bus encoding for low-power DSP systems,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Volume: 9 , Issue: 5 , pages:656 – 660, Oct. 2001.
[3] Y. Aghaghiri and F. Fallah, M. Pedram, “Irredundant address bus encoding for low power,” International Symposium on Low Power Electronics and Design, 6-7, pages:182 – 187, Aug. 2001.
[4] Y. Aghaghiri and F. Fallah, M. Pedram, “ALBORZ: Address Level Bus Power Optimization,” Proceedings. International Symposium on Quality Electronic Design, 18-21, pages:470 – 475 March 2002.
[5] W. Fornaciari and M. Polentarutti and D. Sciuto and C. Silvano, “Power optimization of system-level address buses based on software profiling,” CODES 2000. Proceedings of the Eighth International Workshop on Hardware/Software Codesign, 3-5 , pages:29 – 33, May 2000.
[6] L. Benini and G. De Micheli and E. Macii and D. Sciuto and C. Silvano, “Address bus encoding techniques for system-level power optimization,” Proceedings , Design, Automation and Test in Europe, 23-26, pages:861 – 866, Feb. 1998.
[7] S. Osborne and A.T. Erdogan and T. Arslan and D. Robinson, “Bus encoding architecture for low-power implementation of an AMBA-based SoC platform,” IEE Proceedings- Computers and Digital Techniques, Volume: 149 , Issue: 4 , pages:152 – 156, July 2002.
[8] Chun-Gi Lyuh and Taewhan Kim, “Low power bus encoding with crosstalk delay elimination,” 15th Annual IEEE International ASIC/SOC Conference, 25-28,pages:389 – 393, Sept. 2002.
[9] P.J. Aldworth, “System-on-a-chip bus architecture for embedded applications,” (ICCD '99) International Conference on Computer Design, 10-13 , pages:297 – 298, Oct. 1999.
[10] Youngsoo Shin and Kiyoung Choi, “Narrow bus encoding for low power systems,” Proceedings of the ASP-DAC 2000. Asia and South Pacific , Design Automation Conference, 25-28, pages:217 – 220, Jan. 2000.
[11] M. Madhu and V.S. Murty and V. Kamakoti, “Dynamic coding technique for low-power data bus,” Proceedings IEEE Computer Society Annual Symposium on VLSI. 20-21, pages:252 – 25, Feb. 2003.
[12] Neil H.E Weste and Kamran Eshraghian “Principles of CMOS VLSI Design” Addison Wesley, Second Edition, 1993.
[13] Luca Benin, and Giovanni De Micheli, “System-level Power Optimization of Special Purpose Applications: The Beach Solution” ISLPED, pp.24-29, 1997.
[14] Mircea R. Stan, and Wayne P. Burleson, “Low-Power Encodings for Global Communication in CMOS VLSI” IEEE Trans. VLSI System, pp.444-455, Dec. 1997.
[15] D. Dobberpuhl et al., “A 200 mhz, 64 b, dual issue CMOS microprocessor,” in Proc. Int. Solid-State Circuits Conf., 1992, pp. 106–107.
[16] C. A. Neugebauer and R. O. Carlson, “Comparison of wafer scale integration with VLSI packaging approaches,” IEEE Trans. Comp., Hybrids, and Manufact. Technol., June 1987, pp. 184–189.
[17] A. Shen, A. Ghosh, S. Devadas and K. Keutzer, “On average power dissipation and random pattern testability,” in ICCAD-92, Santa Clara, CA, PP.402-407, Nov. 1992.
[18] A. Park, and R. Maeder, “Codes to Reduce Switching Transients Across VLSII/O Pin,” Computer Architecture News, pp.17.21, sept. 1992.
[19] J. L. Hennessy, and D. A Patterson, “computer architecture: a quantitative approach,” Morgan Kaufmann Publishers, 3nd edition, 2003.
[20] Y. Shin, S. Chae, and K. Choi, “Partial bus-invert for power optimization of system level bus,” ISLPED, pp. 127-129, 1998.
[21] Y. Shin, S. Chae, and K. Choi, “Partial bus-invert coding for power optimization of application-specific systems,” IEEE Trans. on VLSI Systems, Vol. 9, pp. 377-383, April 2001.
[22] C. L. Su, C. Y. Tsui, and A. M. Despain, “ Saving Power in the Control Path of Embedded Processors, ” IEEE Design & Test of Computers, pp.24-30, winter 1994.
[23] L. Benini, G. De Micheli, E. Macii, D. Sciuto, and C. Silvano, “Asymptotic zero-transition activity encoding for address buses in low-power microprocessor-based system,” Six Great Lakes Symposium on VLSI, pp. 77-82, 1997.
[24] L. Benini, G. De Micheli, E. Macii, D. Sciuto, and C. Silvano, “Address bus encoding techniques for system-level power optimization,” Design Automation and Test in Europe, pp. 861-866, Feb. 1998.
[25] Y. Aghaghiri, F. Fallah, and M. Pedram, “EZ Encoding: A class of irredundant low power codes for data address and multiplexed address buses,” Proceedings of the 2002 Design, Automation and Test in Europe Conference and Exhibition (DATE’02), pp. 1102, 2002.
[26] P. E. Landman, and J. M. Rabaey. “Activity-sensitive architectural power analysis,” IEEE Trans. On CAD, pp.571-587, June 1996.