Traders Bet Big on Lisa Su: Why This Chip Titan Is Primed to Siphon Nvidia’s Data Center Dominance - 24/7 Wall St.

This article analyzes how AMD CEO Lisa Su is positioning her company to challenge NVIDIA's market dominance in the data center chip segment. The analysis highlights that as AI and cloud computing dema... Read original →

AMD’s breakthrough in performance-per-watt via Zen 4c and MI300X architectures is eroding NVIDIA’s AI training stronghold. This triggers a cascade: cloud providers like Cloudflare are accelerating adoption of ROCm-compatible stacks to bypass CUDA’s walled garden and cut compute costs. U.S. export controls on advanced chips inflate NVIDIA’s compliance overhead, while AMD’s chiplet-based designs—mixing 7nm and 5nm nodes—enhance supply chain resilience. In response, NVIDIA will likely fast-track Grace-Hopper integration and offer custom IP licensing to lock in hyperscalers. Over the next 18 months, surging inference workloads for large models will favor heterogeneous CPU+GPU systems, enabling AMD to capture over 20% of edge AI data center deployments and dismantle the GPU-only paradigm.