Amazon Taps Qualcomm for Custom AI Data-Center Chips as SEC Filing Ties Warrant Vesting to Up to $60 Billion in Purchases
Amazon and Qualcomm have entered a multigenerational collaboration covering custom AI-inference silicon and optical networking, while an SEC filing reveals an Amazon equity warrant whose vesting milestones are tied to Qualcomm purchases reaching as much as $60 billion.
Why it matters: This moves Qualcomm materially deeper into hyperscale AI infrastructure, with a commercial framework whose warrant milestones extend to as much as $60 billion in purchases.
Qualcomm has secured one of the most consequential hyperscale relationships of its data-center expansion, announcing a multigenerational collaboration with Amazon to develop customized silicon for large-scale artificial-intelligence infrastructure.
The companies will work together on chips aimed initially at AI inference and on high-bandwidth optical connectivity extending to 1.6 terabits per second. The accompanying securities filing makes the economics potentially much larger.
Qualcomm disclosed that it issued an Amazon affiliate a warrant to purchase up to 25 million Qualcomm shares at an exercise price of $161.26 per share. The warrant vests in tranches tied to commercial arrangements, binding purchase orders and actual Amazon purchases of Qualcomm server-chip products, technology, systems and manufacturing services. The filing says those purchase milestones extend to a maximum of $60 billion in payments. That does not mean Amazon has committed to spend $60 billion; it is the maximum commercial-purchase milestone used for warrant vesting.
The relationship centers on two increasingly important bottlenecks in AI infrastructure: compute and connectivity. Qualcomm is expanding beyond smartphones into automotive, PCs, robotics and data-center computing, while Amazon continues to diversify its AI infrastructure across its own Trainium and Graviton silicon and third-party semiconductor partners.
The collaboration gives Qualcomm major hyperscale customer validation and demonstrates how AI infrastructure is becoming an integrated system spanning CPUs, accelerators, networking, optical interconnects and custom silicon.