Mouser Now Stocking NXP Ara240 Discrete Neural Processing Unit, 40 eTOPS Compute Supports Edge LLM
Mouser Electronics is now stocking NXP Semiconductors' new Ara240 discrete neural processing unit (DNPU), which delivers up to 40 eTOPS of compute and supports up to 16GB of LPDDR4 memory, targeting real-time generative AI, industrial automation, autonomous robots and edge applications. The chip can efficiently execute LLMs and VLMs, integrates into edge computing platforms via PCIe Gen4 and USB 3.2 host interfaces, and is accompanied by the Ara SDK to accelerate deployment. Mouser also offers the ARA2-M2-16G-GT M.2 accelerator module based on the Ara240.
Body · AI translation
October 29, 2026 – Mouser Electronics, the authorized global distributor of electronic components and industrial automation products focused on introducing new products, is now stocking the new Ara240 discrete neural processing unit (DNPU) from NXP® Semiconductors. The Ara240 DNPU supports real-time generative AI, industrial automation systems, autonomous robots, smart infrastructure, advanced human-machine interface (HMI) platforms, and edge applications.
The Ara240 DNPU combines balanced compute performance with high off-chip bandwidth to efficiently execute large language models (LLMs) and vision-language models (VLMs), while also offering low latency, low operating costs, and enhanced data privacy. The Ara240 uses an innovative architecture to support advanced multimodal and Transformer-based workloads, delivering up to 40 eTOPS (equivalent trillion operations per second) of compute and supporting up to 16GB of LPDDR4 memory. This capability enables large, complex models to run smoothly directly on edge devices without relying 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 enable easy integration into edge computing platforms, PCs/laptops, and AI 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.
The Ara240 DNPU is accompanied by the feature-rich Ara SDK, which accelerates AI application development and simplifies the deployment of various AI workloads on the Ara240 DNPU and its modules.
The Ara SDK features a scalable AI compiler that supports diverse and evolving model architectures, including convolutional neural networks (CNNs), fully connected neural networks, vision Transformers (ViTs), and highly complex language models such as LLMs and VLMs. The SDK provides efficient dataflow optimization, flexible quantization methods, and support for multiple data types, ensuring high performance and low latency for edge AI applications.
In addition, Mouser also offers the ARA2-M2-16G-GT, an AI accelerator module based on the Ara240 DNPU. This module uses the M.2 M-Key design specification and supports up to four PCI Express 4.0 lanes in an x4 configuration. It uses the common NGFF 2280 form factor and can easily serve as an auxiliary processing unit for the FRDM i.MX 95 and FRDM i.MX95 Pro development boards with M.2 M-Key PCI Express capability.