Micron’s $22B Memory Deals and Qualcomm’s Dragonfly Chips Challenge Nvidia
NVDA•Micron disclosed $22 billion in customer commitments to lock in high-bandwidth memory supplies for AI data centers, underscoring persistent demand that supports Nvidia’s processor ecosystem. Qualcomm introduced its Dragonfly data-center chip lineup with a Meta partnership on the same day, marking a direct challenge to Nvidia’s dominance in AI compute.
1. Micron’s $22B Memory Commitments
Micron reported that customers have committed $22 billion to secure high-bandwidth memory chips used alongside Nvidia’s AI processors, highlighting a supply-constrained environment. Executives said they see no clear timeline for memory supply to meet surging demand, indicating bottlenecks could persist into late 2028.
2. Qualcomm’s Dragonfly Data-Center Push
Qualcomm unveiled its Dragonfly chip series for data centers and announced a strategic collaboration with Meta Platforms to integrate its AI processors in large-scale cloud deployments. This move represents Qualcomm’s most aggressive challenge yet to Nvidia’s market share in AI compute.
3. Implications for Nvidia’s Ecosystem
The combination of tight memory supply and rising competition signals both support and pressure for Nvidia. While secured HBM supplies reduce a key bottleneck, Qualcomm’s entry into high-performance AI chips could erode Nvidia’s pricing power and platform dominance over time.
