Our new RLCk engine pushes the envelope
in frequency and capacity.
Our new RLCk engine, the driving force behind our product line, sets new standards in high-speed IC design by pushing the limits in both frequency and capacity. Our modeling engine outperforms any other electromagnetic modeling tool currently available in the market.
The new architecture of our modeling engine achieves simultaneously:
The need to model electromagnetic effects from DC up to mm-wave frequencies calls for special handling of layouts. A novel full 3D meshing algorithm segments the conductors’ volume into small cells suitable for accurate modeling of capacitance, inductance and resistance. The engine computes all the Layout Dependent Effects (LDE) before the meshing step.
Helic’s 3D capacitance extraction methodology uses a sophisticated stochastic sampling algorithm based on the Random Walk theory for calculating the electric field along the Gaussian surfaces and corresponding coupling capacitances between arbitrary shaped conductors. The solver calculates with the highest accuracy the distributed 3D electric field, using stochastic sampling and a sophisticated numerical solution of the multi-layer Green’s function. The method does not use any kind of pattern matching look-up tables or averaging and is free of conductor discretization bottlenecks. It scales way better with circuit size than boundary or volume meshing methods and demonstrates the best computing efficiency since Random Walk is an inherently parallel and extremely fast algorithm.
Helic’s unique extraction engine models substrate coupling effects with a distributed RC network. A stochastic Monte Carlo based methodology and a 3D substrate model allows for very fast and accurate extraction of the distributed RC substrate network. The method employs a random-walk algorithm that allows characterization of multiple substrate layers using appropriate Green’s functions without the need of three-dimensional discretization. The parallel nature of both capacitance and substrate modeling algorithms offers scalability and extraction times superior to any other method.
Combines the accuracy of a full-wave electromagnetic modeling engine with the flexibility and interoperability of spice netlist output. Extracted models fully capture inductance and resistance behavior from DC up to mm-wave frequencies. Extremely accurate, capturing all electromagnetic phenomena, including current distributions, skin and proximity effects.