#jetsonorin

RT-2 Takes 847ms Per Action. OpenVLA? 1.2 Seconds. You've trained a vision-language-action...

KENSAT Home-Built CubeSat Running: A home-built 2U CubeSat runs a TinyLlama model on a Jetson Orin Nano and does its AI thinking in orbit before radioing results back to Earth.

My Jetson Orin Nano has 8 GB of memory and Gemma 4 26B-A4B needs 13.3 GiB at Q4. I patched llama.cpp to stream the routed experts off the SSD instead, with logits bit-for-bit identical to the stock path, and recorded the model decoding on device.

The 200ms Frame That Shouldn't Exist My ROS2 depth estimation node was hitting 5 FPS. The...