35 Reasons I'm Still Short Nvidia (NASDAQ:NVDA) - Seeking Alpha

This article by analyst Jeremy Blum critically examines the mounting challenges facing NVIDIA despite its dominant position in the AI chip market. The author highlights several headwinds, including an... Read original →

NVIDIA’s AI chip dominance is entering a phase of structural erosion. Technically, while the Rubin GPU buys time, TSMC’s 3nm capacity constraints and ASML’s EUV delivery delays are throttling its roadmap. More critically, hyperscalers like Microsoft and Google are fast-tracking in-house TPUs and ASICs, building vertically integrated AI stacks that bypass GPU generality. On compliance, U.S. export controls force NVIDIA to sell downgraded chips for China—eroding margins and accelerating adoption of domestic alternatives like Huawei’s Ascend. AMD is exploiting this window, capturing Meta and AWS design wins with MI300X, while geopolitical fragility in Taiwan, China is pushing hyperscalers to diversify supply chains. Over the next 12–24 months, AI capex will shift from scale-at-all-costs to efficiency-driven deployment. With over 80% of revenue tied to just five customers—all actively reducing dependency—the risk of order cliff isn’t speculative; it’s a matter of timing, likely materializing before 2027.