Secure processor cores added to Cortex-M range

October 27, 2016 // By Peter Clarke
Processor intellectual property licensor ARM Ltd. (Cambridge, England) has added two processor cores to its Cortex-M range aimed at microcontroller and embedded applications.

These are the Cortex-M23 and Cortex-M33 which are based on the ARMv8-M architecture, which is the latest architecture upgrade for-real time deterministic embedded processor and adds TrustZone security and stack limits to the previous architecture ARMv7-M. Lead partners on the two cores include Analog Devices, Microchip, NXP, Renesas, Silicon Labs and STMicroelectronics.

Cortex-M33 representation. Source: ARM

The difference between the cores is that the Cortex-M33 is aimed at performance while the Cortex-M23 is focused on energy efficiency.

The Cortex-M33 offers 20 percent performance increase in the same process technology compared to the Cortex-M3 and Cortex-M4 processors, while improving power efficiency. It features configuration options including a coprocessor interface, DSP and floating point computation, with increased performance and efficiency relative to Cortex-M3 and Cortex-M4. The Cortex-M23 adds security for heavily power constrained devices similar to those applications previously addressed by the Cortex-M0+.

Cortex-M33 highlights. Source: ARM.

The processors can run code written for the ARMv6-M and ARMv7-M architectures but provide an integrated TrustZone CryptoCell-312.

Cortex-M23 highlights. Source: ARM.

The Cortex-M33 processor has an in-order 3-stage pipeline. Most instructions complete in two stages, while more complex instructions require three. Some 16-bit instructions are dual-issued to boost performance. The core has two AMBA-5 AHB5 interfaces: C-AHB and S-AHB, which are symmetric and offer identical performance of instruction and data fetches. The Cortex-M23 has a two-stage pipeline.

Related links and articles:

White paper on ARMv8-M

News articles:

ARM takes VR/AR mobile with GPU core

ARM reveals R52 core for ‘highest functional safety standards’

ARM's Bifrost steps up graphics, bridges to machine learning