ARM announces purpose-built chips for the Internet of Things, including a Cortex-M85 microcontroller core for IoT devices, and a Corstone-1000 design for edge AI processing.
ARM Holdings’ latest Cortex-M85 microcontroller design will be able to deliver high-performance computing at the edge. The designer also announces two platforms to accelerate the development of IoT devices. Normally, the Cortex-M line of controllers is aimed at small, low-power devices, like smartwatches, but the Cortex-M85 performs better than the rest of the line. In particular, it improves certain artificial intelligence operations such as voice recognition on edge devices, connected home products and drones in particular. “Developers are preparing the IoT of tomorrow, but they must both meet ever-increasing demands for performance and security and follow less complex development flows,” said Mohamed Awad, vice president IoT, in a press release. & Embedded at the chip designer. The Cortex-M85 is a component of the ARM Total Solutions for IoT program, launched by the British six months ago. These pre-integrated subsystem designs offer a more turnkey experience and require less work from chip designers (Nvidia, Qualcomm, Samsung, Texas Instruments…) to make their designs operational. The program also includes the ARM Virtual Hardware cloud service to test devices based on the platform without physical silicon variants. The service also includes machine learning (ML) models and tools to simplify development and speed product design.
Core to the ARM Total Solutions for IoT program, Corstone is a series of pre-integrated designs that combine ARM CPU core designs with other IP building blocks that accelerate the development of vertical processor designs. As part of this announcement, ARM released two Corstone designs for devices with higher performance needs, one called Corstone-310, whose design is based on the Cortex-M85 core, for speech recognition applications and the other called Corstone-1000 for cloud-native edge devices. Combined with ARM’s Total Solution for Voice Recognition program, the Corstone-310 is aimed at devices like smart speakers, thermostats, drones and industrial robots, which could use voice control.
ARM Total Solutions for IoT combines Corstone SoC reference designs based on Cortex-A and -M with the ARM Virtual Hardware Targets platform for cloud-based virtual testing. (Credit ARM)
Combined with ARM’s Total Solution for Cloud Native Edge Devices offering, the Corstone-1000 is a higher-end version. It is intended for application workloads that require high-performance hardware and run on a full-featured operating system like Linux. The Corstone-1000 is based on the Cortex-A architecture, which is much more efficient than the Cortex-M platform. The Corstone-1000 is also part of the vendor’s SystemReady certification program, which ensures that the CPU and subsystems are fully integrated and ready to use. The Cortex-1000 also supports ARM’s Cassini project, intended to simplify “native” software experiences for developers writing applications running on Cortex-A processors.
Expanded CPU simulation support
Almost always, in order to reduce costs, CPU tests are first carried out in simulators before the actual test silicon is produced. The British provides simulators through its ARM Virtual Hardware (AVH) service. These test platforms allow developers to verify and validate embedded and IoT applications during the full software design cycle without the need for actual physical hardware. They offer a large number of modeling technologies to simplify the development and configuration of processor board farms. Now, the ARM offering supports both Corstone subsystem designs as well as seven Cortex-M cores. Independent software vendors and cloud service providers will be able to test applications with the Cortex-M family. The service also includes ARM-based hardware from partner companies, including those from NXP Semiconductors, ST Microelectronics, and makers of the Raspberry Pi.
We would like to give thanks to the writer of this post for this remarkable content
Dedicated ARM IoT processors Computerworld
Find here our social media accounts as well as other pages related to it.https://www.ai-magazine.com/related-pages/