Recent applications of Monte Carlo methods for semiconductor microdevice simulation

B. Shapo, C. Ball, I. Kizilyalli, U. Ravaioli

Research output: Contribution to journalArticlepeer-review

Abstract

Monte Carlo simulations for semiconductor devices are very time consuming. We have investigated ways to speed up calculations on supercomputers and a method to incorporate overshoot effects in simple drift-diffusion models for submicron devices, using coefficients obtained from Monte Carlo experiments. An Ensemble Monte Carlo algorithm, suitable for self-consistent device simulation, has been vectorized, and in preliminary runs three times faster on a CRAY X/MP 48 supercomputer. The inclusion of overshoot terms in a drift-diffusion simulation for a MESFET structure, shows that the main overshoot effects can be incorporated in a simple model suitable for circuit simulation. Increased problems in stability, however, reduce the efficiency of traditional finite difference schemes and require further refinement in the numerical methods.

Original languageEnglish (US)
Pages (from-to)39-43
Number of pages5
JournalSuperlattices and Microstructures
Volume4
Issue number1
DOIs
StatePublished - 1988

ASJC Scopus subject areas

  • Materials Science(all)
  • Condensed Matter Physics
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Recent applications of Monte Carlo methods for semiconductor microdevice simulation'. Together they form a unique fingerprint.

Cite this