In the world of AI chips the smartest moves sometimes come from rivals rather than allies.
Santa Clara, October 2025. In a rare public comment, the CEO of Nvidia, Jensen Huang, described the recently announced partnership between longtime competitor AMD and OpenAI as “imaginative” and “clever”. According to Huang, AMD’s decision to offer equity and supply-chain commitments to OpenAI represents a bold strategic bet in the accelerating race for artificial-intelligence infrastructure. While the move challenges Nvidia’s dominance, Huang acknowledged the game has changed.
The AMD–OpenAI deal features warrants allowing OpenAI to acquire up to ten percent of AMD in return for future chip deliveries and compute-power commitments. Huang noted that giving away that amount of equity before full production begins was “surprising” yet termed it a smart manoeuvre. His comments highlight how AI-hardware competition now intertwines financial engineering and technology in unprecedented ways.
Industry analysts interpret Huang’s remarks as a signal of shifting alliances within the semiconductor sector. Nvidia remains the dominant supplier of AI accelerators, but the AMD agreement shows the market bifurcating: big tech players are seeking diversified supply chains, and chip makers are seeking broader strategic partnerships. For Huang and Nvidia, acknowledging a rival’s smart move is part tactical de-escalation and part positioning for the next phase of competition.
Huang emphasised that Nvidia’s own collaboration model with OpenAI is very different. He pointed out that Nvidia sells its full stack direct to large-scale AI builders rather than offering equity in exchange for future purchases. That distinction matters because it underscores Nvidia’s aim to remain vendor-centric, whereas the AMD approach signals vendor-plus-equity. Both strategies reflect a broader transformation in how compute power is constructed and monetised.
From Asia to Europe and North America, the ripples are clear. In China the banning of certain Nvidia exports coincided with intensifying domestic chip-development efforts. In Europe, chip-policy debates now reference broader issues of supply-chain sovereignty. And in the Americas, investors view AMD’s victory as proof that long-term hardware commitments are valued at the same level as the silicon itself.
Still, the AMD-OpenAI deal carries risks. The infrastructure required by OpenAI involves large‐scale deployment of next-gen chips and multiyear commitments. Analysts warn that the viability of such agreements depends on execution, not just announcements. Huang’s guarded praise perhaps reflects appreciation for boldness tinged with awareness of the uncertainty ahead.
For Nvidia, the acknowledgement that a rival crafted a clever strategy is noteworthy. It suggests a competitive posture evolving from outright dominance to one of adaptation and defence. As Huang helps steer Nvidia into its second decade of AI leadership, recognising the intelligence in a competitor’s move may be just as important as building the next accelerator.
In summary, the chip war is morphing into a war of strategy: hardware, finance and partnerships now merge. Huang’s remarks offer insight into how the most powerful players view each other—not just as adversaries but as strategic actors in a game that is rewriting the rules of computing itself.
Phoenix24: analysis that transcends power. / Phoenix24 : análisis que trasciende al poder.