LattePanda Unveils Mu Ultra: A Compact Solution for On-Device AI

On September 9, 2026, LattePanda Team introduced the LattePanda Mu Ultra, a micro x86 compute module designed for on-device AI applications. Equipped with Intel Core Ultra 5 226V and Ultra 7 256V processors, it achieves up to 115 TOPS of AI performance. The module's compact size of 69.6 × 60mm makes it suitable for integration into space-constrained devices.

The Mu Ultra runs AI tasks directly on the device, reducing latency and cloud dependence. Tests showed it can generate text at speeds of 18 tokens/s with Qwen3.5-9B and 55 tokens/s with Qwen3.5-2B using INT4 quantization. With 16GB of LPDDR5X-8533 memory, it supports a range of AI workloads from language models to sensor processing.

Furthermore, the module offers compatibility with current x86 systems, easing integration for engineering teams. Its modular design allows upgrades without extensive redesign, supporting existing carrier boards and AI tools like OpenVINO.