Mouser Electronics is now stocking NXP Semiconductors’ new Ara240 discrete neural processing unit (DNPU) from NXP Semiconductors, designed to bring real-time generative AI capabilities to industrial automation, autonomous robotics, smart infrastructure, advanced HMI systems, and edge applications.
The Ara240 combines efficient computing performance with high off-chip bandwidth to run large language models (LLMs) and vision-language models (VLMs) with low latency, reduced operating costs, and improved data privacy. Its architecture supports multimodal and transformer-based AI workloads, delivering up to 40 eTOPS of performance and supporting up to 16 GB of LPDDR4 memory. This allows complex AI models to operate directly at the edge, reducing dependence on expensive cloud computing resources.
With integrated secure boot and a hardware root-of-trust processor, the Ara240 DNPU provides hardened security for industrial, enterprise, and embedded deployment environments. Its PCIe Gen4 and USB 3.2 host interfaces allow straightforward integration into edge compute platforms, PCs/laptops, and AI-enabled embedded systems. Whether designing industrial automation systems, autonomous robots, smart infrastructure, advanced human‑machine interface (HMI) platforms or edge servers, the Ara240 DNPU provides the performance headroom needed to run modern AI workloads directly at the edge.
Complementing the Ara240 DNPU is the full-featured Ara SDK, which accelerates AI application development and simplifies deployment of diverse AI workloads across Ara240-based DNPUs and modules.
The Ara SDK features an extensible AI compiler to support diverse and evolving model architectures, including convolutional and fully connected neural networks (CNNs), vision transformers (ViTs), as well as highly complex language models like LLMs and VLMs. The SDK offers efficient dataflow optimization, flexible quantization methods and support for multiple datatypes, ensuring high-performance and low-latency for edge-AI applications.
The ARA2-M2-16G-GT, also available from Mouser, is an AI accelerator module based on the Ara240 DNPU. It uses the M.2 M-Key design specifications and supports up to four PCI Express 4.0 lanes in x4 configuration. Designed with the commonly used NGFF 2280 form factor, it can be easily used as a companion processing unit with the FRDM i.MX 95 and FRDM i.MX95 Pro development boards with M.2 M-Key PCI Express capability.
