Nvidia GTC Unveils NemoClaw, Robot Olaf and Rubin AI Supercomputing Platform, Backed by

Nvidia used the GTC stage to announce three flagship initiatives that it says will lock the company into the next wave of AI‑driven computing. The first, NemoClaw, is a modular robotic arm built around the company’s new H100 Tensor Core GPUs and the latest version of its CUDA‑accelerated control stack. According to TechCrunch, the device is meant to give developers a “plug‑and‑play” way to prototype AI‑powered manipulation tasks, from warehouse sorting to precision assembly, without having to design custom hardware. Nvidia is positioning NemoClaw as the hardware counterpart to its “OpenClaw strategy,” a phrase Huang coined in his keynote to urge every enterprise to adopt a dedicated AI‑edge actuator platform as part of its broader digital transformation roadmap.

The second reveal, Robot Olaf, is a whimsical, anthropomorphic robot that performed a brief demo before Huang cut its microphone due to a technical glitch. While the robot’s on‑stage mishap drew laughs, the underlying technology is serious: Olaf runs on Nvidia’s latest Jetson AGX Orin modules and showcases the company’s progress in real‑time perception, language understanding, and expressive motion. CNET notes that the robot is intended as a proof‑of‑concept for future deployments in entertainment venues, theme parks, and retail spaces where “every company needs an ‘OpenClaw strategy,’” echoing Huang’s call for AI integration at the physical‑world interface.

The third and most consequential announcement is the Rubin AI supercomputing platform, a turnkey solution that bundles Nvidia’s DGX H100 servers, the new DGX Cloud service, and a suite of software tools for large‑scale model training and inference. Precedence Research, cited by TechCrunch, describes Rubin as “the next generation of AI supercomputing,” capable of delivering petaflop‑scale performance with a focus on energy efficiency and automated workload orchestration. Nvidia says the platform will be available both on‑premises and via its expanding cloud partner ecosystem, giving enterprises the flexibility to scale from a single rack to a global hyperscale deployment without re‑architecting their codebase.

Huang used the same keynote to double‑down on the financial stakes of these launches, projecting $1 trillion in AI‑chip sales by 2027. The trillion‑dollar forecast, reported by TechCrunch, rests on the assumption that the demand for AI acceleration will explode across every sector—from autonomous vehicles and robotics to generative media and scientific research. To back that claim, Nvidia highlighted a pipeline of new OEM agreements, including several undisclosed automotive manufacturers that will embed H100‑based compute modules into next‑generation self‑driving platforms. The company also referenced a surge in enterprise AI adoption, noting that more than 2,000 customers have already signed up for its DGX Cloud service in the past year.

Analysts listening to the GTC keynote, as summarized by the TechCrunch Equity podcast, see the trio of announcements as a coordinated push to lock customers into Nvidia’s end‑to‑end stack. By bundling hardware (NemoClaw), embodied AI (Robot Olaf), and a supercomputing platform (Rubin), Nvidia aims to reduce the friction of moving from prototype to production. The podcast hosts warned that startups will need to align their roadmaps with Nvidia’s roadmap or risk being left behind, especially as the company tightens its licensing terms for the H100 and upcoming Hopper‑based GPUs. The Rubin platform’s pricing model, while not disclosed, is expected to follow a subscription‑plus‑usage structure that mirrors the company’s recent shift toward recurring revenue streams.

The broader industry reaction underscores the high stakes of Nvidia’s trillion‑dollar bet. CNET’s coverage points out that competitors such as AMD and Intel are accelerating their own AI‑focused product lines, but Nvidia’s dominance in the GPU market and its deep software ecosystem give it a “foundational” edge, according to the TechCrunch report. Meanwhile, venture‑backed AI startups are scrambling to secure early access to the new hardware, hoping to differentiate their models with the raw performance of the H100‑powered Rubin platform. If Nvidia can convert the announced pipeline into actual shipments, the $1 trillion sales target could become a realistic milestone, reshaping the economics of AI hardware for the next half‑decade.