Self-Driving Cars Stall Under Prohibitive Power Demands

Intelligent Clock Distribution Is Essential for Autonomy

Automotive applications contain a multitude of SoCs for a variety of functions.

As Advanced Driver-Assistance Systems (ADAS) push toward level 5 fully autonomous vehicles, system power requirements increase drastically. This is due to the huge computational requirements for sensor fusion and AI-based image processing and decision making. The heterogeneous nature of autonomous automotive platforms combined with huge power demands makes it essential to have an efficient clock distribution network. The Maestro intelligent clocking solutions can automatically sense and compensate for on-chip variation and skew in the SoC clock path. This allows faster SoC clock rates, more rapid timing closure, and lower clock network power, peak current, and switching noise. Maestro can dramatically improve SoC performance and power efficiency in autonomous applications.

With architectures very similar to automotive platforms, UAS (Unmanned Aircraft Systems) or drones add the complexity of three-dimensional travel and complex battery requirements. This includes high energy to weight ratio, high discharge rates (five-minute to two-hour missions), resilience to shock and vibration, and fuel gauging to indicate remaining mission time.

Maestro can simplify system designs and dramatically improve performance and power efficiency in automotive, autonomous, and other safety-critical applications. It enables an all-digital workflow, unlike traditional analog clocking solutions, that allows full test coverage (scan enabled) to meet ISO 26262 requirements and to comply with other safety standards.