Advanced methods in circuit simulation

The microchip market continues to expand rapidly, driven by increasing demand for electronic communication devices, autonomous transport, connected objects, big data and the Internet of Things.

For decades, this demand has led to a race to miniaturize transistors, which has had the beneficial effect of increasing computing power while reducing size, power consumption and unit costs, in a virtuous circle of investment and high profitability. However, the pursuit of this aggressive downscaling strategy, and in particular the descent into nanometric dimensions, is revealing new challenges and pushing design and verification tools to reinvent themselves. The size of the circuits to be simulated, given the increase in the number of components and interconnections, is becoming critical. The surface densities of dissipated power are becoming so great, that their impact on device operation and lifetime must be effectively taken into account. This is true for all electronics applications, but remains a sensitive issue, particularly in transport, medical and military applications, where reliability is a prime consideration. Similarly, sensitivity to variations in manufacturing processes is becoming so important that taking it into account is a key factor in reducing production costs. Another effect of the descent into nanometric dimensions is the drastic increase in production costs, given the complexity of the manufacturing processes themselves, driving manufacturers to invest massively in production units, with the aim of reducing unit costs.

Against this backdrop, this article shows how the AMS/RF simulator, which is used in both the design and verification phases of circuit design, is a decisive factor in reducing time-to-market, ensuring robustness and reliability, and improving production yields. This article therefore looks at methods for accelerating AMS/RF simulator performance, then at statistical methods for analyzing process variability, and finally at methods for achieving reliable, robust circuit design. The effectiveness of these methods will in part determine whether we can continue the miniaturization process, which boosts performance while reducing power consumption and unit costs.

A glossary and table of acronyms are provided at the end of the article.