Robotics manufacturers face a critical tradeoff between on-device and cloud computing. Datacenter GPUs offer vastly superior compute economics for large foundation models, but wireless latency and jitter continue to force many developers to run robot brains locally.