Cortex-A76AE was launched earlier this year as part of its Automotive Enhanced IP to showcase Arm’s commitment to speed up deployment of completely autonomous vehicles for OEMs. Now, they are further expanding this portfolio of Enhanced Automotive IPs with Arm Cortex-A65AE that was previously known as Helios-AE.
Fig. 1: Representational Image Of Cortex-A65AE
Recent studies made by AAA states that 73% drivers and 63% adults in US do not feel well about sharing driving space with completely autonomous vehicles. Better acceptance of ADAS (Advanced Driver Assistance Systems) will only be possible when humans feel safe and secure about such systems. It is not easy for such manufacturers to win consumer’s trust and demands better solutions capable of balancing safety with innovation while being scalable, deployable and completely ready for mass production.
Features
The Arm Cortex-A65AE has been designed for better processing of varied sensor data streams being produced for next-generation vehicles and to enable innovative driver experiences. The new IP will be delivering higher multithreading ability with integrated safety via innovative Split-Lock technology.
For getting higher autonomy levels, it comes with increased number of sensors keeping better check on vehicle surroundings including radar, LiDAR, and cameras. This helps in more efficient computation of safety requirements for autonomous vehicles. The multiple sensor input enable car drivers to keep check of surroundings, monitor what’s happening around and strategize possible paths and convey commands to actuators over well-defined path.
Since such large amount of data is being collected, it is crucial for system to have higher data throughput capability as part of heterogeneous processing mix needed for allows autonomous and ASAS applications. It is important to keep safety at the core of such systems. The Cortex-A65AE is suitable for taking care of better throughput requirement for collecting sensor data. It can also be used for the lock-step mode in connection to accelerators like computer vision or ML for processing data more efficiently.
Filed Under: News