US Chip Ban Backfires: Huawei's Ascend 910B Leads in China

In a significant turn of events for the global semiconductor industry, US export controls designed to curtail China’s technological ascent are having an unforeseen consequence: they are cementing Huawei’s position as the nation’s domestic AI champion. This outcome validates a stark warning from Nvidia CEO Jensen Huang, who argued that making the US supply chain unreliable for strategic assets is a more powerful deterrent than any technical specification. For China’s tech giants, the risk of interruption has made supply chain sovereignty the new priority. This strategic shift is reshaping the competitive landscape, with Nvidia’s own financial disclosures showing that China and other restricted regions historically contributed 20-25% of its Data Center revenue, a market segment now facing profound transformation due to these geopolitical pressures.
Key Points
• Nvidia CEO Jensen Huang’s analysis demonstrates that supply chain unreliability, created by US export controls, makes American components an unacceptable risk for critical national infrastructure.
• Huawei’s Ascend 910B chip, manufactured by SMIC, has emerged as the primary domestic alternative, with performance measurements showing approximately 80% of Nvidia’s A100 GPU capabilities in specific AI tasks.
• Nvidia’s latest compliant chip for China, the H20, faces documented weak demand, with market reports confirming it is being priced below Huawei’s Ascend 910B, signaling a definitive market shift toward domestic solutions.
• The measurable effect of these controls is the acceleration of a bifurcated global tech ecosystem, a development that CSIS research confirms is “slowing, not stopping” China’s AI progress while fostering state-backed competition.
When Supply Chain Security Trumps Performance
The core of the issue has shifted from a debate over teraflops to one of strategic vulnerability. Nvidia CEO Jensen Huang articulated this during the DealBook Summit, framing the conversation around the logic of national security. His argument posits that no sovereign nation would build its critical military or infrastructure systems on components that could be cut off at any moment by foreign policy decisions. This perspective on China AI supply chain sovereignty is not about chasing the absolute highest performance, but about securing guaranteed, long-term access to essential technology.
This strategic calculus directly explains China’s massive investment in its domestic semiconductor industry. The unreliability of the American supply chain, deliberately engineered by US export controls, provides the ultimate justification for self-sufficiency. As comments from the Nvidia CEO on China’s self-sufficiency drive imply, this dynamic also subtly defends Nvidia’s continued business in China. By framing military use as strategically illogical, Nvidia can argue its compliant chips serve commercial needs, aligning with the stated US policy goal of targeting military applications while minimizing broader economic harm.

Sanctions as Kingmaker: Huawei’s Rise
The primary beneficiary of this policy has been Huawei. US sanctions have inadvertently created a protected, high-demand market for its homegrown AI hardware. The company’s Ascend 910B chip is now the leading viable alternative to Nvidia’s hardware in China, with tech giants like Baidu and Tencent placing significant orders to mitigate their reliance on foreign suppliers.
Manufactured by China’s top foundry, SMIC, on what is believed to be a 7-nanometer process, the Ascend 910B demonstrates competitive capabilities. Industry analysis indicates that in certain AI training workloads, the 910B achieves about 80% of the performance of Nvidia’s A100. This makes it a powerful and, most importantly, domestically sourced option. The US chip ban backfiring on Huawei is evident as the October 2023 blacklisting of other Chinese GPU firms like Biren Technology and Moore Threads effectively cleared the competitive field, leaving Huawei to dominate the domestic high-end AI chip market.
Compliance Costs: Nvidia’s Market Erosion
Nvidia’s attempts to navigate the complex web of regulations highlight the commercial erosion it now faces. After the October 2023 rules banned its A800 and H800 chips, the company developed a new, significantly less powerful trio of compliant chips, led by the H20. However, the market’s reception has been tepid at best.
Pricing has become a clear signal of market dynamics. Reports from the South China Morning Post indicate the H20 is being offered at a discount to Huawei’s Ascend 910B, a direct reflection of weak demand and the H20’s inferior performance specifications. While Chinese cloud companies are testing the H20, their order volumes for Huawei’s chip are substantially larger. This market behavior is reflected in broader economic data; in 2023, China’s imports of integrated circuits fell by over 15% in value for the second consecutive year, a trend that points toward growing domestic replacement capabilities.
Digital Iron Curtain: The Tech Ecosystem Divide
The long-term fallout from this geopolitical chess match extends beyond any single company. Analysts largely agree that while the controls are succeeding in slowing China’s access to the absolute cutting edge, they are also galvanizing its domestic industry. A report from the Center for Strategic and International Studies (CSIS) concludes that the policy is “unlikely to stop China’s progress indefinitely” and will impose long-term costs on US firms.

The most significant risk, as the Jensen Huang warning on US export controls has repeatedly highlighted, is the bifurcation of the global tech world into two competing, incompatible ecosystems. This fragmentation threatens to slow global innovation by creating redundant standards and fractured supply chains, a trend that analysis from the Rhodium Group highlights as a key outcome of China’s push for technological self-sufficiency. While Nvidia maintains over 90% of the global AI accelerator market, its dominance within China is being actively dismantled. The US government is now reportedly considering new restrictions on Chinese access to American AI cloud services, demonstrating the escalating and perhaps intractable nature of creating a complete technological blockade.
Geopolitical Paradox: Creating What You Fear
The US strategy to contain China’s AI ambitions through export controls has produced a complex and ironic outcome. Rather than crippling the industry, the policy has validated the push for self-sufficiency and inadvertently crowned a national champion in Huawei. The market now prioritizes supply chain security over peak performance, a fundamental shift that is eroding Nvidia’s once-unassailable position within China. The result is an accelerated fracturing of the global tech ecosystem, a long-term consequence that may prove more damaging than the short-term technological gap it creates. As these parallel systems develop, what will be the ultimate cost of technological sovereignty for global innovation?
Tags
Read More From AI Buzz

Vector DB Market Shifts: Qdrant, Chroma Challenge Milvus
The vector database market is splitting in two. On one side: enterprise-grade distributed systems built for billion-vector scale. On the other: developer-first tools designed so that spinning up semantic search is as easy as pip install. This month’s data makes clear which side developers are choosing — and the answer should concern anyone who bet […]

Anyscale Ray Adoption Trends Point to a New AI Standard
Ray just hit 49.1 million PyPI downloads in a single month — and it’s growing at 25.6% month-over-month. That’s not the headline. The headline is what that growth rate looks like next to the competition. According to data tracked on the AI-Buzz dashboard , Ray’s adoption velocity is more than double that of Weaviate (+11.4%) […]
