Nvidia CEO: Memory Chip Shortage Could Last for Years - صحيفة مال

NVIDIA CEO Jensen Huang warns that the current memory chip shortage could persist for several years, highlighting critical supply chain vulnerabilities in the semiconductor industry. This announcement... Read original →

Jensen Huang’s warning signals a structural mismatch between AI compute demand and advanced packaging capacity. HBM3E/HBM4’s reliance on TSV and CoWoS processes has effectively locked SK Hynix and TSMC into a de facto alliance. NVIDIA’s partnership with SK Hynix is less about collaboration than securing bandwidth-critical memory for Blackwell and Rubin architectures. U.S. CHIPS Act restrictions on tech coordination between Taiwan, China and South Korea are fragmenting the supply chain, raising compliance overhead. In response, AMD and Micron are fast-tracking HBM integration, while Intel leverages its OSAT assets to push Foveros Direct as an alternative. Over the next 18 months, HBM—not GPU cores—will become the true bottleneck in AI hardware competition, reshaping investor valuation models toward firms with advanced packaging capabilities.