NextFin News - On January 20, 2026, during a high-stakes broadcast of CNBC’s "Mad Money," financial commentator Jim Cramer issued a definitive call for investors to increase their exposure to Nvidia (NVDA). Speaking from the New York Stock Exchange, Cramer emphasized that despite the stock's historic multi-year rally, the fundamental shift toward artificial intelligence (AI) is still in its early innings. He argued that Nvidia remains the "gold standard" for the modern compute economy, particularly as enterprises transition from experimental AI models to full-scale production environments.
The timing of this recommendation is critical. It comes just days after U.S. President Trump’s administration implemented a 25% tariff on certain advanced AI chips, including Nvidia’s H200 and AMD’s MI325X, as part of a broader strategy to incentivize domestic semiconductor manufacturing. While the market initially reacted with volatility to these trade policies, Cramer dismissed the long-term impact on Nvidia, noting that the company’s pricing power and the sheer necessity of its hardware for hyperscalers like Meta and Microsoft outweigh the temporary friction of import duties. According to CNBC, Cramer believes that the record-breaking quarterly results from Taiwan Semiconductor Manufacturing Company (TSMC) earlier this month have "renewed investors' faith" in the sustainability of the AI trade, proving that demand for high-end silicon remains insatiable.
Nvidia’s dominance is no longer just a story of hardware; it is increasingly a story of software and architectural continuity. While competitors like Advanced Micro Devices (AMD) have made significant strides with their MI350 and MI400 series, Nvidia’s CUDA software platform continues to serve as a formidable moat. In 2026, the industry is looking beyond the current Blackwell generation toward the upcoming "Vera Rubin" architecture. This next-generation platform is expected to integrate HBM4 memory and advanced agentic software capabilities, further distancing Nvidia from its rivals in terms of performance-per-watt and ease of deployment. Cramer noted that while AMD is a "credible second source," Nvidia is the architect of the entire ecosystem.
From a financial perspective, the data supports this bullish outlook. As of mid-January 2026, Nvidia maintains a gross margin north of 75%, a figure that remains the envy of the semiconductor world. While its price-to-sales multiple remains high—sitting at approximately 34x compared to Intel’s 4x—the growth trajectory justifies the premium. The shift from "training" large language models to "inference" (running the models) has opened a massive new revenue stream. Analysts estimate that the inference market will account for over 60% of AI chip demand by 2027, and Nvidia’s specialized Tensor Cores are uniquely positioned to capture this volume.
Geopolitical factors also play a nuanced role in this analysis. U.S. President Trump’s administration has signaled that 100% tariffs could be on the table for companies that do not move manufacturing to U.S. soil. However, Nvidia’s strategic partnership with TSMC for its Arizona-based fabs provides a hedge against these regulatory risks. Furthermore, the "Sovereign AI" movement—where nations like Japan and Singapore build their own domestic AI infrastructure—has created a global bidding war for Nvidia’s H-series and B-series chips. According to The Business Times, even with the new 25% duties, the impact on global hubs like Singapore has been minimal, as the demand for these specific high-performance benchmarks remains inelastic.
Looking ahead, the primary risk to Nvidia is not a lack of demand, but rather supply chain concentration and the potential for a "digestion period" among big tech spenders. However, Cramer argues that the rise of "agentic AI"—software that can autonomously perform complex tasks—requires a level of compute density that only Nvidia can currently provide at scale. As we move deeper into 2026, the focus will shift from who has the best chip to who has the best integrated system. With its NVLink networking and full-rack solutions, Nvidia is no longer just a component maker; it is the primary utility provider for the intelligence age. For investors, Cramer’s message is clear: the cost of not owning Nvidia remains higher than the risk of buying it at these levels.
Explore more exclusive insights at nextfin.ai.
