Arm, the company behind the semiconductor design for much of the IoT (and all of the worlds’ smartphones) has announced three projects to help make developing for the internet of things faster, easier, and potentially more secure. It’s announcing all three under the grand name of Arm Total Solutions for IoT.
First, Arm will provide a set of pre-defined hardware capabilities common to the IoT chip manufacturers called Arm Corstone. Then, Arm is launching an abstraction layer for its IP based on Cornerstone that will reside in the cloud. This lets developers access the chip virtually before it exists in hardware. It will also launch Project Centuari, which will let developers use cloud-native software across all of many variations of IoT chips.
This is the second time Arm has tried to solve the many challenges associated with fragmentation in the IoT world. The first time it tried was back in 2014 with the mbed OS. Back then, Arm was trying to unify the many real-time operating systems in use across microcontrollers used in the IoT. That didn’t take.
Because there are so many different types of workloads for the IoT (from the massive ML processing on autonomous vehicles to the needs of a battery-sipping remote sensor) there are myriad operating systems and chip configurations. A developer might be writing code for an embedded RTOS used on a microcontroller or a full application processor running Linux. If they are working with an MCU there are many different physical configurations and options that can also complicate things.
This time around ARM is attacking several problems and has also used some of its experience in building ecosystems for cloud computing (which is also more heterogeneous from a software perspective than Arm’s traditional world of smartphones) to bring development practices such as DevOps and Continuous Integration/Continuous Deployment from the world of cloud computing to the development of IoT products and software.
But first, Arm has to tackle the timeline problem. When Arm announces a new processor design, other chip companies take the IP and use it to build the silicon. That process can take a year or two before the chips are actually in the hands of people who want to build new products. From there, the device makers have to build their own software and hardware and test it. That process can take from three to five years, which, in a world where software is the defining feature and value of a product, is way too long.
Arm’s Corstone plan and the virtual hardware target are trying to solve that delay. With Corstone chip companies such as NXP, Qualcomm or others can build basic processors (Corstone is not for companies that use an Arm architectural license to customize the Arm IP) that contain pre-defined hardware and firmware for common IoT needs such as basic machine learning, over-the-air-software updates, and some security. Arm has been testing this idea with some of its silicon customers for the last few years.
Today it will add this as a regular option for its Arm Cortex M55 processor and Arm Ethos U55 microNPU. Also today Arm will provide a virtual version of the Corstone package in the cloud so developers can start building software for products using the chips when the architecture is released. Mohamed Awad, VP of IoT and Embedded at Arm, says this will cut down on about two years of design time from the announcement of a new Arm architecture to the release of products using that architecture.
This is pretty cool, but the really interesting news is Project Centauri, which is Arm’s attempt to make it much easier for all kinds of developers to build for the IoT. Arm has taken a lesson from its Project Cassini work building a software ecosystem for Arm chips on servers and tried to bring a common software ecosystem to the IoT running on Arm architecture. Cassini runs on Arm’s A-class of processors that run smartphones, servers, and computers, while Centauri will focus on the M-class of microcontrollers.
This will build on work and standards that Arm has announced earlier this year tied to security and CI/CD software development standards common to those familiar with DevOps. Basically, it will make building security features into IoT devices easier as well as allow for development to happen faster and with more standardization.
All of this is good news for IoT customers because it will result in better, more secure products that come to market faster. It will make developers’ lives easier if they choose to use it. I believe that many of them will because some of the features such as Arm’s Corstone are the only things that Arm and its chip customers can provide.
The news also pulls into question the business models for many of the IoT platform companies out there trying to build a common OS layer for multiple IoT devices. For example, companies such as Mason or Foundries both are banking on the underlying complexity of developing hardware and software for IoT to attract customers. My hope is that many of them will look at the base features that Project Centauri offers and build on top of those.
This is big news for the IoT and will open it up to more developers as well as build a faster, more secure foundation for those already building products.