Qualcomm Wants to Bring AI Data Center Power to Your Smartphone - 24/7 Wall St.

Qualcomm is pushing to bring data center-level AI computing power into smartphones and other edge devices, shifting AI processing from the cloud to local hardware. As generative AI grows, reliance on ... Read original →

Qualcomm’s HBC architecture aims to shift AI inference from cloud to edge devices, triggering a cascade across the semiconductor stack: LPDDR and TSV-based stacking gain strategic importance, undermining HBM’s relevance in edge applications and pressuring Micron and SK hynix to accelerate low-power, high-bandwidth memory development. Geopolitically, tightening U.S. export controls on advanced packaging tools could inflate 3D integration costs, especially for supply chains reliant on foundries in Taiwan, China. In response, NVIDIA may double down on Grace CPU integration for lightweight edge AI, while AMD leverages Xilinx IP for heterogeneous compute. If thermal constraints are resolved within 12–24 months, HBC could enable on-device generative AI in smartphones and vehicles, catalyzing a structural pivot from cloud-centric to edge-native AI infrastructure.