In a surprising twist to its China strategy, NVIDIA’s RTX6000D AI chip, designed specifically to comply with U.S. export restrictions, has met with lukewarm demand from major Chinese tech firms, according to multiple sources familiar with procurement discussions. The chip, which began shipping in early September 2025, was expected to help NVIDIA regain lost ground in the Chinese AI hardware market—but early signs suggest the rollout is falling short of expectations.
The RTX6000D, a scaled-down inference GPU based on NVIDIA’s Blackwell architecture, was engineered to meet U.S. export controls while still offering competitive performance for AI workloads. However, sources say that leading Chinese companies including Alibaba, Tencent, and ByteDance have opted not to place significant orders, citing concerns over cost-effectiveness and performance limitations.
Price vs. Performance: A Tough Sell
Priced at approximately ¥50,000 (around $7,000), the RTX6000D is seen by many buyers as overpriced for its capabilities, especially when compared to banned high-end models like the RTX5090, which remain available through grey market channels at nearly half the cost. Testing of sample units reportedly revealed that the RTX6000D lags behind the RTX5090 in key benchmarks, further dampening enthusiasm among enterprise customers.
This mismatch between price and performance has led some analysts to question the chip’s commercial viability. While NVIDIA had projected production volumes of 1.5 to 2 million units for the second half of 2025, actual demand may fall significantly short of those targets.
Strategic Context: Navigating Export Controls
The RTX6000D is part of a trio of downgraded chips—alongside the H20 and the upcoming B30A—developed by NVIDIA to comply with tightening U.S. export restrictions aimed at curbing China’s access to advanced AI hardware. These chips are designed to skirt performance thresholds imposed by Washington while still offering viable solutions for AI inference and training.
However, the rollout of these chips has been complicated by regulatory uncertainty and shifting buyer preferences. While the H20 received clearance for export in July, shipments have yet to resume, and Chinese firms are reportedly holding out for the more powerful B30A, which has not yet been approved.
The tepid response to the RTX6000D underscores the delicate balance NVIDIA must strike between regulatory compliance and market competitiveness. It also highlights the growing sophistication of Chinese buyers, who are increasingly willing to explore domestic alternatives or grey market imports when foreign offerings fail to deliver value.
Market Implications and Industry Reaction
The RTX6000D’s slow uptake could have ripple effects across NVIDIA’s supply chain and revenue forecasts. The company has already absorbed $4.5 billion in inventory impairments and $2.5 billion in unrealized revenue due to previous export bans. A weak performance in China—once a market that contributed over 13% of NVIDIA’s annual revenue—could further strain its financial outlook.
Meanwhile, Chinese chipmakers like Huawei and Cambricon continue to push forward with domestic AI accelerators, though they face challenges in fabrication, bandwidth, and software compatibility. Despite these hurdles, Beijing’s push for self-reliance in semiconductor technology is gaining momentum, and the RTX6000D’s reception may accelerate that shift.
What’s Next for NVIDIA in China?
NVIDIA remains committed to the Chinese market, with CEO Jensen Huang making multiple visits this year to present compliant GPUs to key partners. The company is betting that its B30A chip, which offers higher performance while staying within export limits, will generate stronger demand once approved.
In the meantime, NVIDIA’s spokesperson acknowledged the competitive landscape, stating: “The market is competitive—we offer the best products we can”. Whether that will be enough to win back Chinese buyers remains to be seen.
Sources: Yahoo Finance, Techovedas, NotebookCheck
