Impact of extrinsic and intrinsic parameter fluctuations on CMOS circuit performance

The yield of CMOS logic circuits satisfying a specific high performance requirement is demonstrated to be significantly influenced by the magnitude of critical-path delay deviations due to both extrinsic and intrinsic parameter fluctuations. To evaluate the impact of these parameter fluctuations, a static CMOS critical-path delay distribution is calculated from rigorously derived device and circuit models that enable projections for future technology generations. Two possible options are explored to attain a desired yield: (1) reduce performance by operating at a lower clock frequency; and (2) increase the supply voltage and, consequently, power dissipation, to satisfy the nominal critical-path delay. For the 50-nm technology generation, the delay and power dissipation increases are 12%-29% and 22%-6%, respectively, for extrinsic parameter standard deviations ranging from (a) 5% for effective channel length and 0% for gate oxide thickness and channel doping concentration to (b) 10% for effective channel length and 5% for gate oxide thickness and channel doping concentration. Combining both extrinsic and intrinsic fluctuations, the delay and power dissipation increase to 18%-32% and 31%-53%, respectively, thus demonstrating the significance of including the random dopant placement effect in future CMOS logic designs.

[1]  Gordon E. Moore,et al.  Progress in digital integrated electronics , 1975 .

[2]  A. R. Newton,et al.  Alpha-power law MOSFET model and its applications to CMOS inverter delay and other formulas , 1990 .

[3]  Takayasu Sakurai,et al.  Delay analysis of series-connected MOSFET circuits , 1991 .

[4]  Anantha P. Chandrakasan,et al.  Low-power CMOS digital design , 1992 .

[5]  D. Burnett,et al.  Implications of fundamental threshold voltage variations for high-density SRAM and logic circuits , 1994, Proceedings of 1994 VLSI Technology Symposium.

[6]  James D. Meindl,et al.  Low power microelectronics: retrospect and prospect , 1995, Proc. IEEE.

[7]  The impact of intra-die device parameter variations on path delays and on the design for yield of low voltage digital circuits , 1996, ISLPED.

[8]  Scott K. Reynolds,et al.  Supply and threshold voltage optimization for low power design , 1997, Proceedings of 1997 International Symposium on Low Power Electronics and Design.

[9]  James D. Meindl,et al.  A stochastic wire length distribution for gigascale integration (GSI) , 1997, Proceedings of CICC 97 - Custom Integrated Circuits Conference.

[10]  Vivek De,et al.  Intrinsic MOSFET parameter fluctuations due to random dopant placement , 1997, IEEE Trans. Very Large Scale Integr. Syst..

[11]  Jeffrey A. Davis,et al.  A stochastic wire-length distribution for gigascale integration (GSI). II. Applications to clock frequency, power dissipation, and chip size estimation , 1998 .

[12]  Azeez J. Bhavnagarwala,et al.  Minimum supply voltage for bulk Si CMOS GSI , 1998, Proceedings. 1998 International Symposium on Low Power Electronics and Design (IEEE Cat. No.98TH8379).

[13]  R. Allmon,et al.  High-performance microprocessor design , 1998, IEEE J. Solid State Circuits.

[14]  James D. Meindl,et al.  A LOW POWER TRANSREGIONAL MOSFET MODEL CMOS GIGASCALE INTEGRATION (GSI) * FOR COMPLETE POWER-DELAY ANALYSIS OF , 1998 .

[15]  S. Thompson MOS Scaling: Transistor Challenges for the 21st Century , 1998 .

[16]  Hiroshi Iwai,et al.  CMOS technology-year 2010 and beyond , 1999, IEEE J. Solid State Circuits.