NXP Ara240 Discrete Neural Processing Unit (DNPU)
The Ara240 Discrete Neural Processing Unit (DNPU) enables real-time generative AI, large language models (LLMs), and vision language models (VLMs) execution on AI-enabled compute and embedded systems, delivering low-latency, lower operational costs, and enhanced data privacy.
Features
AI Frameworks Supported
- TensorFlow
- PyTorch
- ONNX
AI Model Architectures Supported
- Convolutional Neural Networks (CNNs)
- Transformer Models
- Large Language Models (LLMs)
- Vision Language Models (VLMs)
- Vision Language Actions (VLAs)
Performance
- Up to 40 equivalent tera operations per second (eTOPS)*
Security
- Secure boot
- Root-of-trust processor
Memory Interface
- Up to 16 GB low-power double data rate 4 (LPDDR4)
Operating System Support (Runtime)
- Linux
Host Interface
- 4-lane peripheral component interconnect express (PCIe) Gen4
- USB 3.2 Gen 1
Chip Package
- 17 mm x 17 mm flip-chip ball grid array (FCBGA)
*eTOPS = equivalent TOPS
Fact Sheet
Ara240 DNPU Fact Sheet
High-performance, energy-efficient discrete neural processing units (DPNUs) are programmable to run a wide range of neural networks, including transformers for multi-modal generative AI and large language models at the edge.
Data Sheet
Ara240 DNPU Data Sheet - Commercial
Explore the Ara240 DNPU datasheet, including AI model support, performance specifications, memory architecture, interfaces, security features, and operating conditions for edge AI applications.
Data Sheet
Ara240 DNPU Data Sheet - Industrial
Access detailed specifications for the industrial-grade Ara240 DNPU, including AI model support, interfaces, security features, memory architecture, performance metrics, and operating conditions.