Stocks · News reference
Broadcom's custom AI chips and Micron's memory boom offer different risks. Which semiconductor stock has the better risk-reward?
Reference Details
Poisar stores the source's headline, summary and market tags for reference. Reporting and full article text remain with the original publisher.
Keep reading
Despite concerns about AI infrastructure sustainability, semiconductor stocks present buying opportunities as cloud computing companies show strong demand for AI chips. Five stocks are recommended: Nvidia (dominant in AI GPUs), Broadcom (custom AI chips), AMD (inference and agentic AI), Cerebras Systems (high-risk/high-reward inference chips), and TSMC (foundry leader).
The SOX index is on track to snap an 8-day losing streak, its longest since October 2008. Nasdaq 100 stocks added nearly $500 billion.
Monday’s regular session has nine sectors lower and two higher, with growth sectors occupying the top two positions. The leaders are separated rather than clustered, with no single gap clearly standing apart from the other breaks across the table.
Sandisk jumped 7.56% and Micron gained 5.13% as investors bought the pullback in tech and memory shares. Get full market analysis
Micron stock rebounded 4% on Monday after falling nearly 15% the previous week due to competition concerns from Chinese chipmaker ChangXin Memory Technologies. The recovery was driven by Dell CEO Michael Dell's prediction of worsening memory shortages in 2027, which would support higher memory prices and profit margins. Wall Street analysts from Mizuho and KeyBanc also issued bullish outlooks on memory stocks, citing strong demand for agentic AI and expected NAND price increases throughout 2027.
The article argues that Taiwan Semiconductor Manufacturing (TSMC) offers better long-term AI investment potential than Nvidia and Micron through 2030. While Nvidia faces competition from custom chips developed by hyperscalers and Micron faces cyclical memory market pressures, TSMC benefits from its foundry position serving all chip architectures. TSMC's Q3 earnings showed 41% YoY revenue growth with 66.4% gross margins, positioning it to capitalize on the broader evolution of AI workloads beyond just training models.