Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Fitting standard cell performance to generalized Lambda distributions

: Lange, A.; Haase, J.; Mau, H.T.


Atienza, D. ; Ecole Polytechnique Federale de Lausanne -EPFL-; Association for Computing Machinery -ACM-, Special Interest Group on Design Automation -SIGDA-:
GLSVLSI 2011, 21st Edition of the Great Lakes Symposium on VLSI. Proceedings : May 2-4, 2011, Lausanne, Switzerland
New York: ACM, 2011
ISBN: 978-1-4503-0667-6
Great Lakes Symposium on VLSI (GLSVLSI) <21, 2011, Lausanne>
Conference Paper
Fraunhofer IIS, Institutsteil Entwurfsautomatisierung (EAS) ()
performance variability; statistical cell characterization; generalized lambda distribution

It is well known that random fluctuations in integrated circuit manufacturing introduce variations in circuit performance. While a lot of effort has been spent on circuit variability, fitting performance parameter distributions has not been extensively examined. Our work analyzes whether the Generalized Lambda Distribution suits approximating circuit performance characteristics. We focus on statistical standard cell characterization as an important step towards statistical gate-level and system-level analyses. Our results show that the Generalized Lambda Distribution is not applicable to raw leakage power data. However, timing data and dynamic power consumption may be approximated well. The high characterization effort has to be overcome to achieve industrial application.