Chip designer Arm Holdings has their latest microcontroller design for IoT presented, which enables high-end computing at the edge. Two platforms are also new to the program, with the help of which the development time of IoT devices is to be shortened.
Typically, the Cortex-M line of controllers are used in small, low-power devices like smartwatches. The new Cortex-M85 is an exception here, as it is significantly more powerful than the rest of the range. It was developed to improve AI applications such as speech recognition in edge devices, such as smart home products or drones.
“Developers are the driving force behind the future of IoT, but they face an ever-increasing demand for higher performance, more security and less complex development workflows,” said Mohamed Awad, Arm’s vice president of IoT and Embedded, in a statement . As the Internet of Things ramps up, it is our responsibility to create more opportunities for IoT innovation and scale by continually raising the bar for performance, easier development, and software reuse.
The Cortex-M85 is part of ARM’s Total Solutions for IoT program, which started six months ago. It consists of pre-integrated subsystem designs that provide a turnkey solution and reduce the workload for chip designers to get their designs up and running. The program also includes the Arm Virtual Hardware Cloud Service for testing Arm-based devices without the need for many variations of physical silicon. Also included are machine learning (ML) models and tools to simplify development and accelerate product design.
At the heart of the Arm Total Solutions for IoT program is Corstone, a collection of pre-integrated designs that combine Arm’s core CPU designs with other IP building blocks for the rapid development and build-out of vertical processor designs. As part of the announcement, Arm introduced two new Corstone designs for devices with higher performance requirements:
Corstone-310 is part of Arm’s Total Solution for Voice Recognition. It targets devices like smart speakers, thermostats, drones, and factory robots, all of which can be controlled by voice. The Corstone 310 design uses the Cortex-M85 core.
Of the Corstone-1000 turn is a high-end solution that Arm calls a Total Solution for Cloud Native Edge Devices. The Corstone-1000 is intended for application-class workloads that require powerful hardware and run on a full operating system such as Linux. For this reason, Corstone-1000 is built on the Cortex-A architecture, which is significantly more powerful than Cortex-M. Corstone-1000 is part of Arm’s SystemReady certification program, which promises the CPU and subsystems are fully integrated and working right out of the box. Cortex-1000 also supports Arm’s Project Cassini, which aims to make cloud-native software easier for developers writing applications for Cortex-A-based processors.
To keep costs down, CPU tests are almost always performed in simulators first, before creating test silicon. Arm offers simulators for this through its Arm Virtual Hardware (AVH) service. AVH provides test platforms for developers to verify and validate embedded and IoT applications throughout the software design cycle without the need for physical hardware. Several modeling technologies are offered to eliminate or reduce the complexity of building and configuring board farms.
Now Arm has expanded support for the two new Corstone subsystem designs, as well as seven Cortex-M cores. This enables independent software vendors and cloud service providers to test applications with the Cortex-M series. In addition, the service is now also suitable for Arm-based hardware from partner companies such as NXP Semiconductors and ST Microelectronics as well as the manufacturers of the Raspberry Pi.