Electro-thermal-stress coupled simulation for Power-MOS device design

• Electro-thermal model: A more accurate field model for the electrical potentials and for the temperature; compact models for the individual channels or channel segments and a compact thermal model for the substrate.
• Analyses: Accurate determination of the thermal sources and sinks and boundaries; high-end processing techniques for finding hot spots and critical segments; perform stress calculations; determine efficiently ageing; upgrade the tools to deal with power transistors with 100 million mesh nodes or more.
• Include variability and reliability constraints (parameter variations). Derive model order reduction methods (MOR) for capturing the input/output relations of the power transistor; reduced models for the individual field equations require the other field variables as inputs/parameters, thus, if they are entered as parameters, parametric MOR methods are needed to preserve these quantities as symbolic variables in the reduced models
• Electro-thermal-stress: Use the available temperature, current density and stress data to make first estimations or predictions of device, bonding life times, electro-migration limitations and some thermal induced failures like passivation cracks

Transceiver designs at high carrier frequencies and baseband waveforms such as OFDM (Orthogonal Frequency Division Multiplex)

• Transient solvers based on the multirate envelope method in conjunction with spline/wavelet bases for an optimal signal representation. Enhancements for EM-heat simulations.
• Improved co-simulation or even a holistic/monolithic circuit-EM-heat simulation approach for accurately predicting the signal waveforms in the presence of crosstalk, substrate coupling, mismatch, etc.
• Improved yield analysis based on statistical tools for reliability from uncertainty quantification.
• Efficiently predict ageing effects: lifetime models for electro-migration, thermal induced failures, the construction of accurate probability distributions or probability density functions