Sources: Nvidia plans to launch a new AI chip priced between $6,500 and $8,000, well below the $10,000-$12,000 the H20 sold for, after US export curbs
Nvidia (NVDA.O) will launch a new artificial intelligence chipset for China at a significantly lower price than its recently restricted H20 model …
Context & Ripple Effects
Nvidia’s China-specific product path has repeatedly been shaped by U.S. export constraints: it introduced the modified H800 for Chinese customers, then moved to the H20, whose pre-orders were priced at $12,000 to $15,000. The new reported price band marks a sharper affordability shift after restrictions hit the H20.
The company had also been reported to be developing a China-compliant B20 with Inspur. A lower-priced successor suggests Nvidia is continuing to tailor products and distribution for a market where compliance can rapidly alter what its existing lineup can sell.
First-order effects
- Chinese buyers that could no longer obtain the H20 would have a cheaper Nvidia option if the planned chip launches, priced at $6,500 to $8,000 rather than the H20’s reported $10,000 to $12,000 range.
- Nvidia would trade some per-chip pricing for a route back into China’s AI-compute market, extending the localized-product strategy behind its earlier modified H800 for Chinese customers.
Second-order effects
- A lower Nvidia price point raises the competitive bar for domestic alternatives, including products that had been compared favorably with the H20 on some specifications when H20 pre-orders opened.
- Chinese server partners and enterprise buyers would need to reassess system configurations and procurement plans around the new compliant chip, rather than treating the restricted H20 as the reference product.
Third-order effects
- If export controls continue to displace Nvidia’s latest China-eligible products, the market is likely to segment into region-specific AI hardware tiers, with design, pricing, and channel partnerships optimized around regulatory ceilings.
- That segmentation could make software compatibility, system integration, and cost per deployed workload more important competitive levers than a single global flagship specification.
The trend: Export controls are pushing AI-chip vendors toward region-specific, compliance-bounded product portfolios that compete as much on deployable cost and ecosystem fit as on peak performance.