Is Nvidia’s AI Dominance At Risk?

For years, Nvidia has been considered untouchable in the AI chip industry. But on January 27th, DeepSeek, a Chinese firm, claimed to have trained a competitive AI model for under $6 million, sending Nvidia’s stock tumbling by nearly $600 billion—the largest single-day loss in U.S. stock market history. Could Nvidia’s dominance start to crumble?

The DeepSeek Disruption

DeepSeek’s breakthrough challenges a core assumption in AI development: that success requires an immense investment in GPUs. Companies like Microsoft, Meta, Elon Musk’s xAI, and OpenAI (through its Stargate venture) have poured billions into GPU-powered data centers, assuming that more processing power means better AI. However, if DeepSeek’s efficiency claims hold, the AI landscape may be shifting dramatically. According to an insider at a major Nvidia customer, DeepSeek’s success suggests that AI training and operations could be far less hardware-intensive than previously thought—a potentially devastating revelation for Nvidia, whose high-end chips boast gross margins exceeding 90%.

Rising Competition

Nvidia’s GPUs remain essential for training large language models (LLMs), but competition is intensifying in inference—the process of running AI models. New players like Groq are developing chips designed for ultra-fast inference, while cloud giants Amazon, Google, and Microsoft are creating custom silicon to reduce dependence on Nvidia. "Markets hate single-source supplier constraints," noted an executive at one of these cloud providers.

Additionally, the latest AI advances, including DeepSeek’s model and OpenAI’s next-generation systems, emphasise “test-time compute,” or increased inference workloads (a type of machine learning and AI workload that use a trained model to make predictions or solve tasks, designed to be fast, efficient, and scalable). This shift could favor alternative chipmakers, challenging Nvidia’s dominance.

The AI Investment Hype

Since ChatGPT ignited the AI boom in late 2022, Nvidia has been seen as the supplier benefiting most from AI’s rapid expansion. But the market is beginning to question whether this growth is sustainable. DeepSeek’s success raises doubts about the necessity of Nvidia’s high-priced chips, and some investors worry that AI’s returns remain speculative rather than proven.

Yet, Nvidia still has its believers. Some argue that DeepSeek may have exaggerated its efficiency to boost China’s technological standing. Others suspect that DeepSeek’s model was trained by analysing outputs from AI systems developed with Nvidia’s top-tier chips.

Nvidia’s Next Move

Despite growing challenges, Nvidia remains a formidable force. Its proprietary CUDA software and networking infrastructure make it more than just a chip company—it’s a full-system provider, reinforcing its leadership.

Moreover, CEO Jensen Huang isn’t standing still. He’s already eyeing the next frontier: "physical AI," or foundational models that interact with the real world. His vision includes AI-powered robots, autonomous vehicles, and beyond. While Nvidia’s stock took a historic hit, it rebounded by over 8% the next day, signaling that investors still see long-term potential.

Nvidia may be under siege, but it would be premature to count it out. The AI arms race is far from over.