NXP Semiconductors has introduced the i.MX RT700 crossover MCU, designed to bring AI capabilities to edge devices while maintaining ultra-low power consumption. This MCU addresses the growing need for AI at the edge in devices like wearables, smart home systems, and medical equipment, where energy efficiency and performance are critical.
i.MX RT700.
The i.MX RT700’s multicore architecture includes an Arm Cortex-M33 core, running at up to 325 MHz, and a second Cortex-M33 core optimized for always-on, low-power tasks such as presence detection and sensor fusion. It also incorporates the Cadence Tensilica HiFi 4 DSP for high-performance audio and voice processing. With up to 7.5 MB of on-chip SRAM, the MCU can execute complex tasks without relying on external memory, enhancing efficiency and reducing latency.
Block diagram of the i.MX RT7000 crossover MCU.
At the heart of its AI capability is NXP’s eIQ Neutron NPU, which accelerates AI workloads by up to 172 times, allowing the MCU to handle real-time machine learning models like image classification, anomaly detection, and voice recognition without the need for cloud computing. This reduces latency and conserves power, making the MCU ideal for battery-powered devices.
The i.MX RT700 also features NXP’s Energy Flex architecture, which supports heterogeneous domain computing. This enables dynamic power management, significantly reducing power consumption in active and sleep modes by 30% and 70%, respectively, compared to earlier crossover MCUs.
The i.MX RT7000 improves power efficiency over its predecessors.
In addition to its performance and power efficiency, the i.MX RT700 includes robust security features through EdgeLock Secure Enclave, offering secure boot, memory encryption, and device authentication. The MCU is now available for manufacturers seeking to build AI-enabled edge devices without compromising battery life or performance.
