Automotive innovations in the field of ADAS and Autonomous Driving, but also in the field of Domain Controllers, Infotainment and Powertrain Systems claim for more processing power within automotive ECUs. The migration to multicore systems is a logical step, e.g. when thinking about AUTOSAR Adaptive applications like sensor fusion of radar, lidar, camera etc. or deep learning algorithms and computer vision. But also when thinking about AUTOSAR Classic based control applications.
High resolution sensor data has to be processed and intelligent decision making algorithms have to be executed, both meeting real-time requirements and additional safety constraints. Furthermore, legacy code applications have to be migrated into parallel applications and the whole software project has to be partitioned and distributed on automotive multicore platforms.
With SLX, the interaction of tasks and runnables can be revealed as well as the program flow and data dependencies. Additionally, beneficial opportunities for parallelization are disclosed. Based on this information, parallelization of runnable implementation as well as runnable to task mapping and scheduling can be optimized.
To increase safety and efficiency, today’s aircrafts are becoming increasingly intelligent – if manned or unmanned. Several data inputs from communication systems and on-board sensors are being processed in real-time. But the implementation of these navigation algorithms in a safety critical environment is still a challenge. With SLX, software implementers can meet these challenging system requirements as to its state-of-the-art compiler technology and performance estimation capabilities.
Next generation mobile networks like 5G offer radically faster connections and significantly more simultaneous users. Highly dynamic workloads can be handled under tight, real-time constraints and thus increase reliability.
This is enabled by mobile base stations that use hundreds of processor cores. But baseband algorithms must be massively parallelized and optimized for heterogeneous cores in order to fulfill all timing and power requirements. This is where base station makers are facing the problem of optimizing their LTE Advanced Pro or 5G code for in-house or of-the-shelf SOCs, as finding further parallelization opportunities and software distribution require huge manual efforts. With SLX automated programming solutions the most challenging system requirements can be met as to its state-of-the-art compiler technology and full heterogeneity awareness.
Self-flying drones or autonomous robots need to understand the world around them and safely navigate in critical situations. This requires a high level of visual intelligence that is enabled by super fast image processing. But the implementation of these algorithms on a device, where every last drop of battery counts, is still a challenge. With SLX, software implementers can meet these challenging system requirements as to its state-of-the-art compiler technology and full heterogeneity awareness.