N1-665 Edge AI System Launched for Autonomous Robotics
The new hardware enables local execution of vision and language models for autonomous systems.
Updated on Oct. 7, 2026 in Artificial Intelligence
Live Poll
Do you believe integrating edge AI into daily technology will improve overall device efficiency?
The N1-665 edge AI system has been released as a platform for autonomous mobile robots and security infrastructure. The device runs large language and vision models locally with a power profile under 20 W.
Why it matters
By enabling local model execution, the N1-665 reduces reliance on cloud connectivity for real-time decision-making in robotics and smart-city applications. This shift prioritizes low-latency processing for time-sensitive tasks in autonomous environments.
The N1-665 features eight Arm Cortex-A78AE cores, 64 GB of DRAM, and 512 GB of flash memory. The system supports 12 simultaneous 1080p30 video streams using quad gigabit multimedia serial links.
The players
Arm
A semiconductor and software design company that licenses the architecture for the A78AE processor cores used in the system.
The details
The N1-665 utilizes a dedicated neural vector processor—a circuit optimized for the matrix math used in neural networks—to handle LLaVa-OneVision, Llama, and Gemma models. Data intake is managed through quad gigabit multimedia serial links, specialized physical-layer interfaces for high-bandwidth video data. These inputs are processed by the integrated decoder, allowing the system to run complex computer vision and reasoning tasks on-device without offloading data to a remote server.
Timeline
- 2026-10-07
The N1-665 edge AI system specifications were formally announced.
The Tech Race
The launch of the N1-665 directly competes with established edge-computing solutions like the NVIDIA Jetson Orin platform. By focusing on local execution for multimodal models, this system challenges existing market leaders in the autonomous robotics and smart-city infrastructure space.
Developers building autonomous robots can now utilize the Cooper Pro platform to integrate onboard language and vision reasoning. Users should expect to see this hardware deployed first in smart-city security systems where low-latency local processing is critical.
The takeaway
The N1-665 represents a move toward high-performance local AI inference in resource-constrained environments. Monitor upcoming benchmarks for the Cooper Pro platform to see how it performs against established edge inference standards.
Further reading
For broader trends in hardware-accelerated machine learning, explore our section on Artificial Intelligence.
Source note: This article includes information reported by Electronic Design.
Live Poll
Do you believe integrating edge AI into daily technology will improve overall device efficiency?
