NVIDIA Deepens Korean AI Startup Investments at MSIT Forum

At the MSIT‑NVIDIA AI Startup Partnership Forum in Seoul, NVIDIA announced a new round of planned investments in four Korean‑focused startups that span the full AI stack.

The ministry of science and ICT joined NVIDIA executives to meet more than 20 companies working on foundation models, physical AI, AI infrastructure and AI semiconductors. NVIDIA confirmed an investment in Point2 Technology, an interconnect specialist that uses radio‑frequency links to move data between chips and racks, offering a scalable alternative to copper and optical cables for next‑generation AI factories.

Additional commitments target Motif Technologies, which builds open‑weight foundation models for autonomous agents; RLWRLD, a physical‑AI venture developing robotics foundation models for dexterous humanoid and industrial robots; and Upstage, an enterprise AI firm delivering Solar language models and document‑AI tools for business workflow automation.

Why it matters for GPU / AI infrastructure

Point2’s RF interconnect directly addresses the bandwidth and latency bottlenecks that limit GPU cluster scale‑up, a critical concern for cloud providers like AiGpu that operate large‑scale AI factories. Meanwhile, advances in agentic models and robotics foundations increase demand for high‑performance compute, reinforcing the need for flexible, high‑throughput GPU infrastructure.