SDV performance documentation

Huxelerate Software Defined Vehicle (SDV) performance is a support decision tool to:

  1. Correctly size the vehicle E/E architecture by optimizing the number of on-board ECUs and their capabilities

  2. Optimize software by design to meet timing, resources and E/E architecture requirements.

The platform leverages on Huxelerate proprietary technology that allows to obtain performance estimates (execution time, resource usage, optimization opportunities) at the single runnable function granularity without the need of executing the software on the target hardware, considerably saving development, testing and optimization time.

To provide the estimates of performance, Huxelerate platform provides a library of architectural models that modelize the architectural features of a specific processor.

The platform supports multiple development steps, from architecture and communication network definition, to the optimization of the single runnable function that executes on a specific core of a specific processor of an ECU. Thanks to the possibility to create multiple scenarios, it follows software and architecture evolutions, allowing to perform different impact analyses and simulations, while keeping under control timing and performance requirements.

The platform allows to define your vehicle computing architecture in terms of:

  • CPU processors inside the ECU

  • communication channels

The estimator can:

  • Generate Software-In-the-Loop (SIL) wrappers for local simulation of your modules generated via Simulink.

  • Provide implementation for I/O interfaces of the autosar runnable, used to provide input data (either randomly generated or loaded from a CSV file), and output data, dumped to a CSV file during the execution.

  • Estimate actual performance of your software on your target architectures with information on your runnable function and the functions executed within it.

  • Estimate the occupied resources of your functions provided a target execution period.

  • Guide you through the optimization process to potentially improve the performance of your software.

Advantages

The insights provided by the tool allows you to:

  • Solve timing issues in advance by understanding actual performance of your modules before testing on hardware

  • Automatic verification of timing constraints

  • Optimally map variables among available memories analyzing the impact on the overall execution time

  • Optimize your code without wasting computing resources