"Genetically engineered" nanostructure devices

Materials Research Society Symposium - Proceedings

Volumn 551, Issue , 1999, Pages 149-154

  Klimeck, Gerhard ^a   Salazar Lazaro, Carlos H ^a   Stoica, Adrian ^a   Cwik, Thomas ^a  

^a CALIFORNIA INSTITUTE OF TECHNOLOGY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

COMPUTER AIDED ANALYSIS; COMPUTER AIDED DESIGN; COMPUTER SIMULATION; ELECTRON TRANSPORT PROPERTIES; GENETIC ALGORITHMS; HETEROJUNCTIONS; QUANTUM THEORY;

QUANTUM DEVICE DESIGN; SOFTWARE PACKAGE NANOELECTRONIC MODELING TOOL (NEMO); SOFTWARE PACKAGE PGAPACK;

NANOTECHNOLOGY;

EID: 0033354562     PISSN: 02729172     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Article

References (14)

1. NEMO, Nanoelectronic Modeling, in http://www.raytheon.com/rtis/nemo/.
2. R. Lake et al., J. Appl. Phys., 81(12), 7845 (1997).
3. R. Lake et al., phys. stat. sol. (b), 204, 354 (1997).
4. G. Klimeck et al., Appl. Phys. Lett., 67(17), 2539 (1995).
5. G. Klimeck et al., IEEE DRC, 1997: p. 92.
6. R. C. Bowen et al., J. Appl. Phys., 81, 3207 (1997).
7. A. C. Seabaugh, Texas Instruments, private communication, 1997.
8. The actual CPU time needed for a single I-V simulation depends strongly on the choice of material systems, bandstructure models, temperature scattering models, and bias points. The individual I-V characteristics presented here take about 30 minutes to compute on a single 200MHz R10000 CPU of an SGI Origin.
9. D. Levine, http://wvw-unix.mcs.anl.gov/∼levine/PGAPACK/index.html, Parallel Genetic Algorithm Library.
10. J. N. Schulman, Second Workshop on Characterization, Future Opportunities and Applications of 6.1Å III-V Semiconductors, Aug. 24-26, 1998, Naval Research Laboratory, Washington, DC, http://estd-www.nrl.navy.mil/code6870/code6870.html.
11. H. C. Liu, J. Appl. Phys. 64, 4792 (1988). .
12. H. C. Liu, J. Appl. Phys. 53, 485 (1988).
13. PGAPACK is implemented with MPI where N-1 of N processors are slaves to one master processor. The master takes care of the collection of data from the slaves. In a cluster of 64 CPU's we therefore renew only 63 genes in every generation.
14. The mutation operations may drive a particular parameter outside its original range. Therefore the full parameter space may not be limited.