back

Huawei’s New AI Chip: A Real Threat to NVIDIA? | Vantage with Palki Sharma

Huawei challenges Nvidia's AI chip dominance

In a stunning technological leap, Chinese tech giant Huawei has unveiled its latest AI chip, throwing down the gauntlet to Nvidia's market dominance. This development comes amid escalating US-China tech tensions and represents a potential shift in the global AI chip landscape that few industry observers anticipated would happen so quickly. The stakes couldn't be higher as AI processing power increasingly determines which companies—and nations—will lead the next wave of technological innovation.

Key Points

  • Huawei's new Ascend 910B processor represents China's most significant challenge yet to US semiconductor supremacy, with performance metrics allegedly comparable to Nvidia's H100 chips despite US export controls designed to hamper Chinese AI development.

  • The geopolitical dimensions are profound, as this chip demonstrates China's determination to achieve technological self-sufficiency in advanced computing despite Western restrictions—potentially reshaping global tech supply chains.

  • While impressive, questions remain about Huawei's manufacturing capabilities, SMIC's 7nm process limitations, and whether their production can scale to meet massive demand as AI deployment accelerates globally.

Expert Analysis

The most striking aspect of this development isn't just the chip itself but what it reveals about the effectiveness of export controls as a technology containment strategy. The US has implemented increasingly strict restrictions on semiconductor technology exports to China—particularly for AI applications—yet Huawei still managed to produce a competitive chip. This suggests that technology containment policies may delay but cannot prevent determined nations from developing indigenous capabilities when they prioritize specific technologies as strategic imperatives.

This matters immensely because AI processing infrastructure represents the foundation upon which virtually all future AI applications will be built. Control of this layer of the technology stack translates directly to economic advantage, military capability, and geopolitical leverage. If China can indeed produce competitive AI chips at scale, it fundamentally alters assumptions about America's ability to maintain its technological edge through export controls alone.

Original Value-Adds

The Huawei breakthrough highlights a pattern we've seen repeatedly in tech history: innovation under constraint often produces unexpected results. When Japan faced US memory chip restrictions in the 1980s, it redirected focus toward consumer electronics, eventually dominating that market. Similarly, South Korea's response to certain technology access limitations was to double down on display technology

Recent Videos

May 6, 2026

Hermes Agent Master Class

https://www.youtube.com/watch?v=R3YOGfTBcQg Welcome to the Hermes Agent Master Class — an 11-episode series taking you from zero to fully leveraging every feature of Nous Research's open-source agent. In this first episode, we install Hermes from scratch on a brand new machine with no prior skills or memory, walk through full configuration with OpenRouter, tour the most important CLI and slash commands, and run our first real task: a competitor research report on a custom children's book AI business idea. Every future episode will build on this fresh install so you can see the compounding value of the agent in real time....

Apr 29, 2026

Andrej Karpathy – Outsource your thinking, but you can’t outsource your understanding

https://www.youtube.com/watch?v=96jN2OCOfLs Here's what Andrej Karpathy just figured out that everyone else is still dancing around: we're not in an era of "better models." We're in a different era of computing altogether. And the difference between understanding that and not understanding it is the difference between being a vibe coder and being an agentic engineer. Last October, Karpathy had a realization. AI didn't stop being ChatGPT-adjacent. It fundamentally shifted. Agentic coherent workflows started to actually work. And he's spent the last three months living in side projects, VB coding, exploring what's actually possible. What he found is a framework that explains...

Mar 30, 2026

Andrej Karpathy on the Decade of Agents, the Limits of RL, and Why Education Is His Next Mission

A summary of key takeaways from Andrej Karpathy's conversation with Dwarkesh Patel In a wide-ranging conversation with Dwarkesh Patel, Andrej Karpathy — former head of AI at Tesla, founding member of OpenAI, and creator of some of the most popular AI educational content on the internet — shared his views on where AI is headed, what's still broken, and why he's now pouring his energy into education. Here are the key takeaways. "It's the Decade of Agents, Not the Year of Agents" Karpathy's now-famous quote is a direct pushback on industry hype. Early agents like Claude Code and Codex are...