Nvidia's Jensen Huang Warns DeepSeek Running on Huawei Chips Is a 'Horrible Outcome' For US

Why It Matters

If leading AI models are optimized for non‑American chips, U.S. hardware firms could lose market share and influence over global AI standards, reshaping the competitive landscape.

Key Takeaways

• DeepSeek on Huawei chips threatens US hardware dominance
• Huang urges US to control AI stack across all five layers
• He opposes full market cut‑off, favors international dialogue
• Export‑control analogy to nuclear weapons called illogical
• Open‑source models could shift global standards away from US

Pulse Analysis

The rapid emergence of China’s open‑source model DeepSeek, now being tuned for Huawei’s Ascend processors, signals a potential pivot in the AI hardware hierarchy. By leveraging a domestic chip architecture, Chinese firms could deliver comparable performance to Nvidia‑based solutions, eroding the United States’ long‑standing advantage in AI compute. This shift would not only affect revenue streams for chipmakers but also influence the de facto standards that guide software development, data formats, and interoperability across the global AI ecosystem.

Huang’s remarks on a recent podcast highlighted a growing tension between security‑driven export controls and the need for a collaborative AI environment. While some industry leaders, including Anthropic’s Dario Amodei, advocate for stricter restrictions—citing parallels to nuclear proliferation—Huang dismissed the analogy as “lunacy,” emphasizing that microprocessors and memory are already globally sourced. He warned that scaring developers away from AI work could backfire, depriving the U.S. of talent essential for innovation and leaving a vacuum that China could fill.

To safeguard its leadership, the United States must reinforce dominance across the entire AI stack—from cutting‑edge silicon to foundational software frameworks. This involves accelerating domestic chip R&D, fostering open‑source initiatives that remain aligned with U.S. standards, and cultivating diplomatic channels that promote responsible AI development worldwide. By balancing strategic protection with constructive engagement, the U.S. can mitigate the risk of a fragmented AI landscape while preserving its influence over the next generation of intelligent technologies.