Will Ebiefung, The Motley Fool
4 min read
In This Article:
This has been a brutal year for the U.S. tech industry as a combination of challenges, ranging from political uncertainty to foreign competition, has shaken some of its core assumptions. Globalism is no longer guaranteed, and developing markets like China are transitioning from a source of low-cost labor into a viable threat to U.S. technological dominance.
The next 12 months could be a make-or-break period for Nvidia (NASDAQ: NVDA) as it navigates the evolving macroeconomic landscape. Let’s dig deeper to decide how the company’s shares might perform.
While it is too early to be sure, President Donald Trump’s April 2 tariff announcement may mark the beginning of the end for globalization — a system that facilitated increasingly complex international supply chains. Nvidia is a typical example of how globalized business models work.
As a fabless semiconductor company, Nvidia only designs its chips. Their actual production is outsourced to foreign companies like Taiwan Semiconductor Manufacturing (TSMC), which produces them with the help of EUV machines from the Dutch company ASML.
These international interdependencies could draw unwelcome political attention. But the good news is that Nvidia worked to preempt this exposure with the help of its manufacturing partners.
In late 2024, TSMC opened its first U.S. factory in Arizona, allowing Nvidia to source its cutting-edge Blackwell AI chips domestically. This move could shield the company from potential tariffs, which is essential for its long-term success. Even though semiconductors are currently exempt from this latest round of tariffs, their strategic nature will likely attract attention from this administration and future U.S. lawmakers, so it is advantageous for Nvidia to get ahead of this potential headwind.
Although trade war talk is stealing the show in April, Nvidia’s stock price crash started in January, with the launch of the high-performing Chinese large language model (LLM) DeepSeek V3, which was allegedly developed for just $6 million. While the actual cost is still hotly debated, it is believed that V3 was built using much less advanced H800 chips while still comparing favorably to American LLMs like OpenAI’s ChatGPT, which were built with substantially more expensive hardware.
DeepSeek calls into question the amount of money American AI companies are spending on Nvidia hardware. And while major customers like Meta Platforms continue to plow money into the opportunity, it may be a matter of time before they face shareholder backlash over the amount being spent for minimal financial reward.