Asus has officially unveiled the PE3000N mini PC, a rugged industrial computer built around Nvidia’s new Jetson IGX Thor T5000 module. It’s not aimed at everyday users but at developers working on robots, autonomous machines, and AI-driven infrastructure.
PE3000N mini PC uses Nvidia’s powerful new AI module

Inside, the PE3000N mini PC runs on Nvidia’s Jetson IGX Thor T5000, a module that combines a Blackwell GPU, 14-core ARM processor, and 128 GB of LPDDR5X RAM. Altogether, it pushes an astonishing 2,070 FP4 TFLOPS of AI performance.
That’s enough to run advanced visual language models, LLMs, and real-time decision systems for robotics, industrial automation, and more. Asus built this to handle serious edge-AI workloads, and it shows.
Rugged build for real-world deployments
Asus engineered the PE3000N mini PC to thrive in demanding environments. It features a ruggedized chassis that meets MIL-STD-810H standards and handles operating temps from -20°C to 60°C.
Whether it’s deployed in a smart factory, warehouse, or autonomous vehicle lab, it’s ready to work nonstop.
High bandwidth and modular expansion
The PE3000N supports high-speed sensor input and wide connectivity, including:
- Up to 4x 25 GbE NICs
- Support for 16 GMSL cameras
- Optional expansion stack with PoE, GMSL, CAN, QSFP28
- PTP/PPS for sensor sync
- TPM 2.0 onboard security
- LTE/5G/GNSS for wide-area coverage
Developers can use USB-A, USB-C, HDMI, and standard audio jacks for familiar I/O needs. But the real power lies in its ability to plug into massive, multi-sensor ecosystems.
Built for generative AI and autonomous systems
Thanks to its underlying platform, the PE3000N can run Nvidia Isaac for robotics, Holoscan for medical sensor processing, and Blueprint for edge video AI. It’s not just powerful, it’s optimized for next-gen AI agents in physical space.
PE3000N mini PC arrives in Q1 2026
As of now, pricing remains unknown. But considering Nvidia’s Jetson IGX Thor dev kit starts around €3,000, the full-featured PE3000N will likely land higher.
Still, for robotics labs and AI integrators, this isn’t just another PC, it’s a launchpad for physical AI.

