SANTA CLARA, California, May 3, 2026, 12:01 (PDT)
- Bloomberg data puts about 90% of Nvidia’s production costs with Asian suppliers.
- Money is flowing out of chipmakers and into robotics, autos, and memory stocks linked to Nvidia, as investors push those names higher.
- With U.S. markets shut Sunday, Nvidia shares last changed hands at $198.45, slipping 0.6%.
Nvidia’s focus on “physical AI”—that is, artificial intelligence powering robots, vehicles, and factory systems operating in the real world—has been pulling more attention to the chipmaker, especially as Asian suppliers take on a larger role in its supply chain. Bloomberg data put Asian firms at roughly 90% of Nvidia’s production costs now, a jump from about 65% a year ago. The Business Times
That’s relevant right now, with investors shifting their focus beyond just data-center chips. In the past week, LG Electronics, Nanya Technology, Huizhou Desay SV Automotive, and Pateo Connect Technology Shanghai all saw their shares climb—following reports of new partnerships, product development, or supply-chain connections involving Nvidia.
This latest development throws Nvidia’s supply chain back in the spotlight as U.S. markets get set to reopen Monday. Shares of Nvidia last changed hands at $198.45—a drop of 0.6%—with the company’s market cap sitting near $4.86 trillion, according to market data.
Global tech giants like Nvidia are bound to rely more heavily on Asian supply chains, according to Ling Vey-Sern, managing director at Union Bancaire Privee. Physical AI, he said, is likely to fuel even greater demand for the region’s suppliers already linked to AI chip production.
Nvidia’s ties to Asian hardware players are growing, with SK Hynix and Samsung Electronics among its key partners on the memory side. Now, the company is pushing beyond servers and chips, targeting fields like robotics, autonomous vehicles, and AI-driven manufacturing—domains where hardware, sensors, and simulation tech carry as much weight as the chips themselves.
Last week, LG Electronics said it had recently held talks with Nvidia about potential work together on robotics, AI data centers, and mobility. Madison Huang, Nvidia’s senior director for physical AI platforms, visited LG along with other South Korean firms.
Nvidia CEO Jensen Huang is positioning robotics front and center in the company’s AI push. “Physical AI has arrived,” Huang declared back in March, rolling out fresh collaborations with robotics firms and predicting a shift: more industrial players turning into robotics companies. NVIDIA Newsroom
Broader demand is pulling in more suppliers to the AI expansion, according to Marvin Chen, strategist at Bloomberg Intelligence. That, he notes, might keep “tech-heavy north Asian markets” ahead of the pack. Moneycontrol
Suppliers aren’t seeing any letup. Last week, Samsung reported its chip unit’s quarterly profit soared 49 times higher, and flagged the risk of even tighter supply next year if AI demand keeps climbing. The company also confirmed it’s now selling HBM4 memory chips to power Nvidia’s Vera Rubin platform.
The competitive landscape keeps moving. Nvidia is fighting to maintain its edge over Intel and Advanced Micro Devices—both veterans in supplying CPUs, the workhorse chips behind most computer operations—while Nvidia carved out dominance in GPUs, the chips now powering much of AI.
The worry? This trade’s become a crowded bet, hinging on a handful of key threads: Nvidia demand keeps surging, Asian firms keep landing contracts, and U.S. export rules don’t snarl the flow. Last week, Reuters flagged that Nvidia’s B300 servers were fetching about $1 million apiece in China as tougher restrictions and a clampdown on smuggling squeezed supply. Nvidia, for its part, confirmed the B300 is barred from China and cautioned that “unlawful diversion is a recipe for failure.” Reuters
Right now, Nvidia’s position looks less like that of a lone chipmaker and more like a hardware lynchpin. Over at Gama Asset Management, portfolio manager Rajeev De Mello points to Asia’s technology base as a structural edge, especially as AI’s appetite stretches across semiconductors, components, servers, and the entire hardware ecosystem.
