2 AI Chip Stocks to Buy on the Dip

2 AI Chip Stocks to Buy on the Dip


After a great run over the past couple of years, top artificial intelligence (AI) stocks have stumbled out of the gate so far in 2025. High valuations, concerns about the economy (and demand for semiconductors), and the near-term direction of spending on data centers are contributing to negative sentiment for these stocks right now.

While the semiconductor industry is cyclical, it has grown for decades, and AI is the catalyst that should lift leading chip stocks higher over the next decade. Dell’Oro Group believes spending on data center infrastructure could surpass $1 trillion in five years. The following stocks are well-positioned to profit from this opportunity.

Shares of Nvidia (NASDAQ: NVDA) are down 23% from their recent highs. While the near-term impact of President Donald Trump’s tariff policies is still unknown for Nvidia and the broader semiconductor industry, Nvidia’s lead in graphics processing units (GPUs) provides ample opportunities for long-term growth and shareholder returns.

Nvidia said during its fourth-quarter earnings call in February that demand for its chips needed for AI inferencing is accelerating. Inferencing is where most of the value for chip companies will come over the long term. This is the stage of AI development in which models become smart enough to make predictions on their own from new data, and it requires significantly more computing power.

Nvidia says AI models that can reason at a very high level will require 100 times more compute per task. This is why companies are investing in its new Blackwell computing platform, which provides up to 25 times more throughput for data units, or tokens, while lowering costs compared to older hardware. Revenue from Blackwell hit $11 billion in the fourth quarter and should increase as production ramps this year.

Large cloud service providers, including Amazon, Microsoft, and Alphabet‘s Google, totaled nearly half of Nvidia’s $35 billion in data center revenue last quarter. But there are plenty more tech companies scrambling to get their hands on the company’s powerful GPUs.

xAI is adopting Nvidia’s GB200 chip to power its Grok AI models. Meta Platforms is using Nvidia’s Grace Hopper super chip for running ads on Instagram and Facebook. Meta said it plans to spend between $60 billion to $65 billion in capital expenditures this year to support plans in generative AI and other business operations, and that’s to Nvidia’s benefit.

Overall, Nvidia’s total revenue grew 12% over the previous quarter and 78% year over year to reach $39.3 billion in Q4. Blackwell will be a key driver in pushing Nvidia’s revenue up to approximately $43 billion in Q1, based on company guidance.



Source link