NXP Semiconductors has introduced its new i.MX RT700 crossover MCU family, designed to deliver substantial power savings and advanced AI capabilities for smart Edge devices such as wearables, medical devices, smart home products, and HMI platforms. The i.MX RT700 series combines high performance, extensive integration, and enhanced power efficiency to meet the growing demands of AI-enabled edge computing.
The i.MX RT700 includes up to five powerful cores in a single device, marking the first use of NXP’s eIQ Neutron Neural Processing Unit (NPU) in a crossover MCU. This enables up to 172x AI acceleration while reducing energy consumption per inference by up to 119x. The series also features up to 7.5MB of ultra-low power SRAM and achieves a 30-70% reduction in power usage compared to earlier models.
The increasing demand for compute power and new functionality in AI-enabled Edge devices, many of which are battery-operated, has driven the need for more efficient MCUs. The i.MX RT700 addresses this by offering a low-power multicore architecture, incorporating powerful graphics capabilities, AI hardware acceleration, advanced security, and a sense compute subsystem. This allows companies to create versatile solutions with features such as gesture recognition, voice control, and presence sensing on a single unified platform.
“As pioneers of the crossover MCU, we’re not just advancing products with the i.MX RT700, we’re redefining what’s possible at the Edge,” said Charles Dachs, Senior Vice President and General Manager, Industrial and IoT, NXP. “The i.MX RT700 provides significant enhancements in power consumption and efficiency, enabling breakthroughs that extend battery life and ensure reliability in resource-constrained applications. Additionally, the integration of our eIQ Neutron NPU enables customers to build innovative machine learning applications, improving AI and multi-tasking capabilities of low-power Edge devices.”
The i.MX RT700 series boasts a range of advanced features, including a primary Arm Cortex-M33 core running at 325 MHz and an integrated Cadence Tensilica HiFi 4 DSP for high-demand DSP and audio processing tasks. The inclusion of NXP’s eIQ Neutron NPU, enabled through the eIQ machine learning development environment, facilitates AI acceleration by offloading workloads from the Cortex-M33. This allows the i.MX RT700 to perform tasks like anomaly detection up to 18x faster and image classification up to 172x faster compared to running these tasks on the Cortex-M33 alone.
To optimise power consumption, the i.MX RT700 incorporates several architectural innovations, including adaptive deep sleep techniques and advanced wake/sleep cycles, resulting in a 30-70% improvement in power efficiency. This allows for extended battery life or smaller battery use, offering greater design flexibility for end devices.
The i.MX RT700 also enhances system performance with an integrated high-efficiency DC-DC converter, improved memory management unit, and better support for analogue peripherals. Notably, it is NXP’s first crossover MCU to support the emerging eUSB standard, enabling USB 2.0 interfaces to operate at I/O voltages of 1V or 1.2V instead of the conventional 3.3V.
In terms of security, the i.MX RT700 is built with the EdgeLock Secure Enclave (Core Profile), providing advanced protection for connected devices. It supports secure boot, memory encryption, and device authentication via physically unclonable functions (PuF). This ensures compliance with emerging cybersecurity regulations, including the European Cyber Resilience Act and ETSI 303 645, safeguarding consumer data and ensuring cyber resilience in Edge devices.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
