NXP Launches Crossover MCU Targeting AI-Enabled Edge Applications
NXP Semiconductors has taken a significant leap forward in edge computing with the introduction of the i.MX RT700
crossover MCU. This innovative device combines high-performance AI
acceleration with ultra-low power consumption, making it ideal for a
wide range of applications where battery life and performance are
paramount.
i.MX RT700.
BonChip Electronics, a leading distributor of electronic components, is excited to offer the i.MX RT700 MCU as part of our comprehensive product portfolio. With our commitment to exceptional service and fast delivery, we can help you seamlessly integrate this powerful MCU into your next-generation design.
A Multicore Marvel for Efficiency
The i.MX RT700 boasts a unique multicore architecture featuring up to five compute cores, each meticulously designed for specific tasks. This intelligent allocation optimizes both performance and power efficiency.
- Central Processing Unit (CPU): The heart of the MCU is an Arm Cortex-M33 core clocked at up to 325 MHz, handling real-time tasks and general computing needs.
- Digital Signal Processing (DSP): A Cadence Tensilica HiFi 4 DSP tackles audio and voice applications with exceptional performance.
- Always-On Sensing: A dedicated Cortex-M33 core within the “sense compute subsystem” operates at 250 MHz, ideal for low-power, continuous monitoring tasks like presence detection, gesture recognition, and sensor fusion. Accompanying this core is a HiFi 1 DSP for low-latency, power-efficient audio processing.
Block diagram of the i.MX RT7000 crossover MCU.
Furthermore, the i.MX RT700 integrates a generous 7.5 MB of on-chip, ultra-low-power SRAM with zero wait-state access. This eliminates the need for external memory, allowing the MCU to execute complex AI tasks efficiently.
eIQ Neutron NPU: Unleashing the Power of AI
The i.MX RT700 truly shines with its integrated eIQ Neutron Neural Processing Unit (NPU). This innovative NPU offloads demanding AI workloads from the main Cortex-M33 cores, accelerating performance by up to 172 times depending on the application. Tasks like image classification, anomaly detection, and keyword spotting become significantly faster and more efficient.
This translates to real-time edge AI inference. Devices can now execute machine learning models locally without compromising battery life. Traditionally cloud-based tasks like object recognition and predictive maintenance can be handled directly on the device, reducing latency and reliance on external infrastructure. The eIQ Neutron NPU further impresses with its ability to reduce energy consumption per inference by up to 119 times, making it perfect for power-constrained devices like wearables and portable medical equipment.
Unparalleled Power Efficiency
One of the biggest challenges in edge AI has been balancing performance with power consumption, especially for battery-powered applications. NXP addresses this head-on with the i.MX RT700's impressive low-power features:
- Reduced Power Consumption: Compared to previous generations like the RT500 and RT600 series, the i.MX RT700 boasts a 30% reduction in active power consumption and a staggering 70% improvement in sleep mode efficiency.
- Energy Flex Architecture: This innovative architecture allows for heterogeneous domain computing, dynamically adjusting power consumption based on the processing load.
- Dual Compute Domains: The MCU is divided into two distinct domains – compute and sense. This ensures each subsystem operates within its optimal power envelope. For example, high-performance AI tasks prioritize the compute domain, while the sense domain remains in a low-power state for always-on tasks.
- Advanced Power-Saving Techniques: Deep sleep modes, optimized clock architectures, and low-power cache schemes further minimize power consumption during idle periods. Notably, the MCU can retain up to 5 MB of SRAM during sleep, enabling a quick wake-up without reloading data.
The i.MX RT7000 improves power efficiency over its predecessors.
A Feature-Rich MCU for the Future
The i.MX RT700 goes beyond AI and power efficiency with a range of additional features:
- Integrated DC-DC Converter and Memory Management Unit: These features optimize system performance.
- Support for New eUSB Standard: This allows USB 2.0 interfaces to function at lower voltages (1V or 1.2V) for even greater power savings.
- Robust Security Features: The MCU incorporates EdgeLock Secure Enclave for secure boot, secure updates, memory encryption, and secure data access. It also leverages device authentication with built-in physically unclonable function (PuF) technology.
With its exceptional performance, low-power consumption, and comprehensive feature set, the NXP i.MX RT700 MCU is poised to revolutionize edge AI applications.
Contact BonChip Electronics today to learn