NVIDIA’s Jensen Huang ‘Surprised’ by AMD’s Excitement Toward the MI450 AI Lineup, Implies Competition Is Ramping Up

In a recent interview with CNBC, NVIDIA CEO Jensen Huang commented on AMD’s growing involvement in the AI acceleration market, particularly its ambitious partnership with OpenAI that centers on the upcoming Instinct MI450. For context, AMD recently secured a multi-billion-dollar deal with OpenAI to supply over six gigawatts of computing power through its upcoming Instinct MI450 AI accelerators, a partnership that positions AMD alongside NVIDIA as one of the primary compute providers for OpenAI’s large-scale infrastructure. The agreement was announced just days after OpenAI revealed a separate collaboration with NVIDIA, immediately signaling a more competitive landscape between the two semiconductor giants.

When asked about AMD’s deal, Huang offered what many perceived as both a candid and slightly humorous reaction, saying:

“It’s imaginative, it’s unique, and surprising, considering they were so excited about their next generation product. I’m surprised that they would give away 10% of the company before they even built it. And so anyhow, it’s clever, I guess.”

Huang’s remarks reference AMD’s decision to dedicate a portion of its future compute capacity and partnership stake to OpenAI, even before the MI450 series has officially launched. His statement highlights NVIDIA’s view that such an early move by AMD is bold, if not risky, especially given the still-unreleased status of the hardware line.

The NVIDIA CEO also noted that his company was not aware of AMD’s negotiations with OpenAI at the time the latter signed its deal with NVIDIA, suggesting that AMD’s move caught the industry off guard.

The AMD–OpenAI deal underscores a significant shift in the AI compute market, where NVIDIA has long maintained a dominant position. With AMD now stepping in with its MI450 architecture, and NVIDIA advancing its Vera Rubin AI platform, both companies are preparing for direct competition across multiple tiers of AI workloads, from GPU compute to rack-scale solutions.

AMD CEO Dr. Lisa Su recently stated that this partnership could generate up to $100 billion over the coming years, emphasizing the scale of the opportunity and AMD’s confidence in its next-generation Instinct lineup. Industry analysts believe this deal could reshape the market dynamics by introducing meaningful competition in a space that has, until now, been largely dominated by NVIDIA.

Still, Huang’s comments suggest that NVIDIA is unbothered by the challenge and even welcomes the competition. “It’s clever,” he admitted, while implying that AMD’s aggressive strategy signals a heightened race for leadership in the rapidly growing AI acceleration sector.

For several years, NVIDIA’s CUDA ecosystem, H100 and H200 GPUs, and DGX platforms have formed the backbone of AI computing for major players including Google, Amazon, and Microsoft. However, AMD’s growing traction, combined with partnerships like this one, could lead to a more diversified market.

Healthy competition often drives innovation, and as AI infrastructure demands soar, both companies are expected to push harder on performance, efficiency, and scalability. With AMD’s MI450 and NVIDIA’s Vera Rubin architectures both aiming for late 2025 and early 2026 availability, the battle for AI supremacy is about to intensify.

Do you think AMD’s aggressive approach will finally challenge NVIDIA’s long-held dominance in AI computing, or will NVIDIA maintain its leadership through its mature ecosystem and software stack?