technologyScore 60/100Research
Palmer Luckey: Autonomous drone systems moving toward on-board compute within 2-3 Moore's Law cycles as chip costs drop below communications infrastructure
Palmer Luckey· Anduril Industries· Robotics· 2026-05-26· about Anduril Industries
“We're like two Moore's Law 18 month cycles away from it being much cheaper to have this level of autonomy, which is on andro drones today, like higher end stuff. The chips cost hundreds of dollars today, which is like twice as much as one of these fiber optic rails or three times as much as a crappy radio. You just wait three or four years. It's going to be cheaper than all those things, less jamable, more resilient.”
Why it matters
This signals a near-term shift in the economics of autonomous weapons systems. As compute costs fall below comms infrastructure costs, system architectures will consolidate around on-board processing, reducing reliance on RF/fiber links and creating new demand for embedded AI chips.
Investment implication
Semiconductor suppliers competing in embedded AI and edge compute for defense applications (NVIDIA, Qualcomm, Intel) should see increased TAM in autonomous weapons. Conversely, RF and fiber optics suppliers may face reduced demand in this segment.