Nvidia in Korea: open license, proprietary silicon - the Isaac platform model

Create a watch on this topic Receive upcoming developments if any, with an AI watch configured from this article.

Create my free watch

Nvidia and Hyundai Motor Group announced on June 8, 2026, at the conglomerate's headquarters in Yangjae (Seoul), a new phase of cooperation centered around three axes: software-defined autonomous vehicles (SDV), humanoid robotics through the Boston Dynamics subsidiary, and an AI factory powered by Blackwell GPUs linked to the Saemangeum project. On the same day, a few kilometers away, LG and Nvidia simultaneously announced the M.A.P. partnership - Mobility, AI Infra, Physical AI - on a largely parallel scope. On the same sequence, Jensen Huang signed six agreements: Unitree on June 2, then SK Hynix, Naver, Doosan, Hyundai, and LG during the Korean tour from June 5 to 8. This pattern is not an ordinary commercial tour: the previous Android is analytically close on three aspects - commercial non-exclusivity, permissive licensing, effective hardware dependency - but differs on a crucial point: Google did not manufacture the silicon for Android, Nvidia designs and sells the silicon that runs GR00T.

An open license, a closed silicon

GR00T N1.7, a 3-billion-parameter Vision-Language-Action (VLA) model published as open-weight under an Apache 2.0 license by Nvidia, combines two stages: a Cosmos-Reason2-2B module that translates images and instructions into action intentions, and a 32-layer Diffusion Transformer that converts these intentions into real-time motor commands. The license allows for forking, retraining, and deployment without fees. Inference performance, on the other hand, is tied to the embedded Jetson Thor module, credited by Nvidia with 2,070 FP4 TFLOPS, 128 GB of memory, and 7.5 times the AI power of the previous AGX Orin. The FP4 format—a four-bit numerical representation—is currently only fully utilized on this silicon; the training ecosystem Omniverse, however, is tied to CUDA. The Nvidia developer page does not publish third-party evaluations of GR00T N1.7 on real manufacturing tasks, and the documentation remains focused on simulation demonstrations.

Hyundai, test case with uncertain implementation

Hyundai's stated goal is 30,000 Atlas robots per year by 2028, in a new robotics plant near Savannah, Georgia—the target announced by the company at CES 2026 and not reiterated by Nvidia in its June 8 statement. According to Hyundai's data, over 25,000 of these units would be destined for the company's and Kia's production lines, accounting for 83% of the targeted capacity. Internal deployment is therefore the main market, and it is here that implementation becomes complicated: Axios reports that the company's union has contested the introduction of robots without prior social agreement, and official communication remains silent on the timeline for such an agreement. On the technical side, Boston Dynamics published a blog describing the learning of carrying heavy loads through internal force feedback rather than visual identification—method that makes Atlas's performance less dependent on the visual reasoning of a VLA model than the platform narrative suggests.

The platform's scope: segment, not universal

Three major actors in the humanoid field remain outside Isaac: Tesla with Optimus, Figure AI, and Apptronik develop their own software stacks and are not dependent on either GR00T N1.7 or Jetson Thor. The platform position documented by the Korean sequence therefore applies to a specific segment: the established Asian manufacturer, which combines strong vertical integration and a desire to automate its own production. The previous Android is analytically similar in three aspects—non-exclusive commercialization, permissive licensing, effective hardware dependency—but differs in one: Google did not manufacture Android silicon, Nvidia designs and sells the one that runs GR00T. The readable signal today is limited: Nvidia holds a platform supplier position in the Korean physical AI industry, not in the entire humanoid market.