ADAS solution with enhanced safety enables ASIL-D compliant system construction

ARM has announced ARM DynamIQ technology for multi-core processor systems. DynamIQ will provide multi-cores for devices ranging from platform terminals that require security to the cloud. It can be applied to smartphones, cars, home appliances, and many other connected devices. It will enable artificial intelligence (AI) that utilizes machine learning to produce zettabytes of data between the cloud and devices.



It can deliver AI performance up to 50x higher than the existing Cortex-A73 system with dedicated processor instructions for machine learning and AI. It improves the response speed between the CPU and the dedicated acceleration hardware of the SoC by up to 10x. SoC designers can scale up to eight cores with different performances in a single cluster. This flexibility improves the response speed of machine learning and AI applications. In addition, software switching controls the operating frequency of each processor. It provides ADAS solutions to build safe ASIL-D compliant systems even in the event of a failure.

ARM announced that it has now shipped more than 100 billion ARM-based chips through its partners, 50 billion of which were achieved in just four years from 2013 to 2017, showing the growing need for computing capabilities. ARM said it expects to ship more than 100 billion additional ARM-based chips in 2021.

“There is a growing need to deliver autonomous systems, AI, and mixed reality (MR) that integrates virtual reality (VR) ubiquitously,” said Nandan Nayampally, general manager, compute products, ARM. “We are announcing DynamIQ technology that enables the highest performance without compromising efficiency.”