China Reportedly Blocks Nvidia RTX 5090D V2 as GPU Politics Heat Up
China has reportedly banned Nvidia’s RTX 5090D V2, the China-focused version of its flagship RTX 5090 graphics card, according to a Financial Times report cited by Tom’s Hardware.
The timing is spicy. The reported customs ban apparently landed while Nvidia CEO Jensen Huang was in China, joining President Donald Trump’s state visit after being added to the trip late. If accurate, that is a pretty loud signal from Beijing: even Nvidia’s specially adjusted, export-compliant GPUs may no longer be welcome.
What is the RTX 5090D V2?
The RTX 5090D V2 is not the same as the full-fat RTX 5090 sold in other markets. Nvidia designed it for China to fit within US export control rules, with reduced VRAM and lower memory bandwidth compared to the standard model.
On paper, this is still mainly a high-end consumer GPU for gaming, 3D rendering, animation, and creator workloads. But because China’s AI companies have been cut off from Nvidia’s most powerful Blackwell AI accelerators, even cards like the 5090D V2 can become attractive for smaller AI development, testing, and local compute setups.
That overlap is why this ban matters beyond just PC gaming. A gaming GPU can still be useful for AI work, especially when dedicated data-centre chips are restricted.
China wants local chips to win
The bigger picture here is China’s push to reduce dependence on American hardware. Tom’s Hardware notes that while the Nvidia H200 was approved for export to China in late 2025, Beijing has reportedly not allowed its AI companies to buy those chips.
Instead, the Chinese government wants companies to support domestic silicon. Huawei is one of the biggest names benefiting from that shift, with China’s AI ecosystem being nudged toward homegrown accelerators instead of Nvidia hardware.
For Nvidia, that is the nightmare scenario. The company does not just want to sell chips; it wants Chinese developers to stay inside its CUDA ecosystem. Once AI firms rewrite workflows around local alternatives, winning them back becomes much harder.
Why Malaysia and SEA should care
At first glance, this sounds like a US-China problem only. But GPU politics always finds its way into SEA pricing, availability, and grey-market chaos.
Malaysia’s gaming PC crowd already knows how fast high-end GPU prices can get weird when supply chains shift. If China tightens access to Nvidia cards while demand stays high, some stock may move through unofficial channels, regional distributors, or resale markets. That can affect prices on Shopee, Lazada, and local PC shops, especially for ultra-high-end cards that already cost serious money in RM.
There is also an AI angle. Malaysian startups, universities, content studios, and small labs may not be buying giant H200 clusters, but they do rely on global GPU availability. When big markets fight over supply, smaller SEA buyers usually feel the aftershock through delayed stock, inflated prices, or limited options.
For gamers, the RTX 5090D V2 itself is unlikely to be a mainstream Malaysia card. But the policy drama around it matters because it shapes Nvidia’s product strategy, regional supply, and how aggressive alternatives from China may become over time.
Nvidia is still hoping the door opens
Tom’s Hardware says it has contacted Nvidia for comment. Huang, meanwhile, has remained publicly optimistic, saying on Bloomberg TV that he believes the market will open over time.
Maybe it will. But right now, China appears more interested in proving it can build an AI hardware stack without relying on Nvidia’s best chips — or even Nvidia’s trimmed-down ones.
Source: Tom's Hardware
