Powertrain controller design is among the most challenging problems due to the complexity of the functions that the system has to support, to the safety aspects and to the cost limits imposed by car manufacturers. To compound these difficulties, time-to-market requirements are becoming more and more stringent. Design time, continuously changing specifications and safety considerations have pushed the design more and more towards software implementation of the main functionality. Software has been traditionally designed with very little abstraction in mind thus forcing a tight dependency of the implementation on the particular hardware architecture, e.g., the instruction set of the micro- controller. Software legacy has made the rapid adoption of new technology and IC's almost impossible, stifling innovation. In addition, the absence of a correct abstraction hierarchy made verifying the correctness of the behavior of the system as well as adding new functionality extremely difficult. To cope with this problem, we have developed a new generation of powertrain controller designs that exploit the new function-architecture co-design paradigm and software architecture principles presented elsewhere in the session. Following this paradigm, we have been able to develop function design to a large extent independently of the particular micro-controller architecture selected for the final implementation. This method has allowed us to re-use a large part of the design effort and, at the same time, to leverage IC technology. We have also been able to identify bottlenecks and inefficiencies common to present IC industry offerings and to propose new architectures that have been developed in collaboration with IC manufacturers. This approach has been made easier to follow by the advent of new advanced system design environments embodying the function-architecture co-design paradigm.