Business

4 No-Brainer AI Chip Stocks to Buy Right Now

Advertisements

Spending on artificial intelligence (AI) infrastructure is set to see a big boost this year, which will undoubtedly help a number of semiconductor stocks with ties to AI chips.

For example, the big three cloud computing companies — Amazon, Microsoft, and Alphabet — budgeted a combined $255 billion in growth capital expenditures (capex), largely related to AI infrastructure, this year. Meanwhile, Meta Platforms announced that it will spend up to $65 billion on AI infrastructure. Not to be outdone, a consortium of companies led by Japan’s Softbank and OpenAI committed to spending $500 billion on AI infrastructure in the U.S. over the next few years through Project Stargate.

Let’s look at the AI chip companies set to reap the benefits.

Nvidia (NASDAQ: NVDA) is the leader in graphics processing units (GPUs), which are the main chips used in AI training and inference due to their superior processing speeds. The chips were originally designed to help speed up graphics rendering in video games. But later, the company let customers program the chips for other tasks through its CUDA software program. Today, through its CUDA X platform, which was built on top of CUDA, it offers GPU-accelerated microservices and libraries designed specifically for AI.

CUDA helped give Nvidia customers a superior experience and created a wide moat for the company. This can be seen in its approximately 90% market share in the GPU space. As such, the company is poised to be one of the biggest winners from increased AI infrastructure spending.

In addition, the stock is attractively priced, trading at a forward price-to-earnings (P/E) ratio of 25 times 2025 analysts’ estimates and a price/earnings-to-growth (PEG) of 0.5, with PEGs below 1 considered undervalued.

While Nvidia is the leader in mass merchant GPUs, Broadcom (NASDAQ: AVGO) has become the leader in helping customers design their own custom AI chips. While these application-specific integrated circuits, or ASICs, lack the flexibility of GPUs, they generally have better performance and more efficient power consumption for the specific task for which they were designed.

Broadcom’s first custom AI chip customer was Alphabet, as it helped it design its tensor-processing unit (TPU) called Trillium. Since then, it added other customers that are believed to include Meta Platforms, ByteDance, OpenAI, and most recently, Apple. Broadcom noted that just its top three customers could deploy up to 1 million AI chips in 2027, representing a $60 billion to $90 billion revenue opportunity in 2027. While Nvidia will likely get a nice piece of that revenue with its GPUs, Broadcom also has a huge opportunity as well.

https://s.yimg.com/ny/api/res/1.2/DQ2QTuibPcmC0zyWrQzlRA–/YXBwaWQ9aGlnaGxhbmRlcjt3PTEyMDA7aD02NzU-/https://media.zenfs.com/en/motleyfool.com/0b23eca3e6d4b4f216a1a0a6e888d743

2025-02-23 15:50:00

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button