RTX 3060 MXM GPU Carrier Board Kit for LattePanda Mu (Local AI & 6x 4K Output)

The LattePanda Mu RTX 3060 MXM Carrier Board Kit combines discrete graphics power with an ultra-compact footprint for edge compute deployments. Ideal for makers, researchers, and digital signage engineers, it handles demanding AI inference and multi-display installations where standard desktop PCs cannot fit. Please note that a LattePanda Mu compute module is required and sold separately.

Discrete RTX 3060 edge computing

Equipped with an NVIDIA GeForce RTX 3060 MXM graphics card with 12GB of VRAM, this kit accelerates generative AI, visual-language models, and complex computer vision workloads. It provides sufficient local capacity to execute INT4-quantised large language models up to approximately 9B parameters at the edge.

Six-channel 4K video outputs

Drive high-density digital signage matrices and video walls with up to six independent 4K displays at 60Hz. The output layout combines three HDMI ports and three DisplayPort connections across the compute module and MXM discrete graphics card.

Modular CPU architecture

The carrier board is compatible with LattePanda Mu compute modules powered by either the Intel N100 or Core i3-N305 processor. This modular design allows you to scale CPU performance independently while retaining full discrete GPU acceleration.

Compact enclosure and connectivity

Measuring 159 x 107.5 x 88.5 mm inside its included metal enclosure, the system fits easily into interactive kiosks and behind commercial displays. External I/O includes two USB 3.2 Gen 2 ports, two USB 2.0 ports, Gigabit Ethernet, and internal M.2 expansion slots.

Power configuration and safety guidelines

Never connect the DC power jack and the 8-pin ATX power supply simultaneously, as dual connection causes permanent board damage. Verify your MXM fan voltage rating and adjust the onboard PWR_FAN jumper between 5V and 12V before powering the device.

Downloads & documentation

• Product wiki (DFRobot)