ARM CPU Family Explained (2019)

Back in 2015 I made a post about the lineup of NXP (ex Freescale) development board. This year I am planning to do the same.

But first, I think it is pertinent to talk about the family of Arm Processors and the target applications of each one of them; given that a big part of micro-controller manufacturers out there have at least one group of Arm based MCU.

Arm’s architecture specifications are licensed by companies, who create compliant silicon chips based on them. They claim to be more than 125 billion devices containing Arm-based chips today in the market and, while their architecture prevails in the mobile market, they also feature in many other areas.[1]

Arm architecture specifies a set of rules that dictate how the hardware works when a particular instruction is executed. It is a contract between the hardware and the software, defining how they interact with one another. 

When software is written to conform with Arm specifications, any Arm-based processor or chipset will execute it in the same way. This is the foundation of the portability and compatibility promise, underlying the Arm ecosystem. Our architecture is consistent, compatible and it delivers.

Arm System and Security Architectures provide standardisation and best practice guidance, enabling the ecosystem to reduce cost and accelerate time to market when designing systems.

Through sustained collaboration with our partners, our next generation of architecture will enable designs that push the boundaries of compute.


Arm’s processors are divided by subfamilies; each one of them dedicated to a different kind of performance and applications. The first one we are going to talk about is the Cortex-A Family.

Designed for high performance and power efficiency, they represent most of mobile devices in the last years, as well as automotive computers.

This one is the most diverse group of architectures, some of the most well known processors of this family are the Cortex-A55 and Cortex-A75 used on the latest release of Snapdragon SoC (as of Q1 2019)


Real time processors at their best. Designed to be reliable and accomplish every task under strict constraints. These architectures are mostly used for automotive and industrial design, where timing is critical.


Designed for secure low cost solutions, this is the family of architectures most used for micro-controllers and embedded design.

The Cortex-M0 is the smallest arm processor available and along with the Cortex-M0+ they are the most common architectures used for IoT.


The next three families are part of the latest additions to the arm family; which include design for Edge Computing, Machine Learning and Security.

The first of them is the Neoverse group of processors, designed for infrastructures, edge and cloud computing. They were recently announce, so I guess we are yet to see what can be done with this new offering.

Machine Learning

What else is there to say? Neural Networks, Machine Learning and Artificial Intelligence are the main focus of these architectures. Designed with efficiency in mind.


The Securcore family are Cortex-M devices on security steroids. Arm included a tamper resistant design and counter measures against side channel attacks and fault injections.

[1] Architecture - Arm. Retrieved February 25th, 2019

Leave a Reply

Your email address will not be published. Required fields are marked *