Nvidia announced the Jetson Orin Nano 2, a robotics and edge AI computer that delivers twice the inference performance of its predecessor while cutting power consumption by 40 percent. The chip is scheduled for availability in the first half of 2027.
The Jetson line targets embedded AI applications where full-scale data center hardware is impractical: autonomous robots, industrial vision systems, and edge devices that must run inference locally. The Nano 2 compresses those capabilities into a smaller power envelope, a constraint that defines the entry-level robotics compute market. Halving power draw while doubling inference throughput addresses the core tradeoff that has constrained adoption in battery-powered and thermally limited deployments.
Nvidia did not disclose pricing for the Nano 2. The original Jetson Orin Nano, released in 2023, launched at $199 and became the company's fastest-adopted edge AI accelerator. Jetson boards ship to robotics manufacturers, industrial OEMs, and research labs; Nvidia has not reported quarterly Jetson revenue separately, but the line is part of the broader Data Center segment, which posted $60.9 billion in full-year fiscal 2024 revenue.
The announcement arrives as robotics startups and established manufacturers race to deploy AI-powered systems. Boston Dynamics, Figure AI, and Tesla's Optimus division have all publicly committed to scaling humanoid and task-specific robots over the next two years. Those platforms require onboard inference capacity to operate in environments where network latency or outages would halt operation. A chip that halves power draw while maintaining or exceeding prior performance removes a material barrier to deployment in production fleets.
Nvidia's Jetson roadmap has historically tracked the company's broader GPU architecture iterations with a lag of one to two generations. The Orin line launched in 2022 and has dominated entry-level robotics compute since. The Nano 2 delivers both efficiency and performance gains. Nvidia sees sustained demand in the segment and is not diverting all advanced lithography capacity to data center and consumer GPU markets.
The timing places the Nano 2 roughly 18 months after the original Orin Nano release and aligns with Nvidia's pattern of annual or biennial hardware refreshes in the edge segment. Competitors including Qualcomm, MediaTek, and Intel have released or are planning comparable edge AI chips, but none have achieved Jetson's penetration in robotics developer communities or its integration depth with Nvidia's CUDA software stack.
The 40 percent power reduction matters most in autonomous systems where weight and battery life determine operational range. A robot operating on lithium batteries can now run 67 percent longer on the same charge if power consumption is the limiting factor, or carry proportionally less battery mass for the same mission duration. That efficiency gain has no published competitor equivalent at the Nano 2's inference tier as of August 2026. Whether the performance and efficiency gains translate to higher unit adoption depends on pricing, which Nvidia will disclose closer to availability, and on whether robotics deployments scale fast enough to absorb supply when the chip reaches production in H1 2027.
