In July, most AI stocks entered a correction as the rotation into sectors such as homebuilders and banks gained momentum. Still, any pullbacks in these stocks present a buying opportunity. We are in the early innings of a new technological era that could continue for decades.
Nvidia’s (NASDAQ:NVDA) CEO Jensen Huang calls this ongoing shift to accelerating computing the next industrial revolution. Bloomberg Intelligence research shows that generative AI will grow at a 42% compounded annual growth rate over the next 10 years. As AI proliferates through industries such as education, gaming and life sciences, it will spur an investment cycle in AI training and inference.
These three AI stocks are at the forefront of the generative AI infrastructure buildout. One is a key supplier enabling the buildout of AI data centers. The other two provide the semiconductors and networking infrastructure needed to accelerate computing for generative AI training and inference.
Vertiv (VRT)
As cloud service providers (CSPs) build out AI data centers to support large language model training and inference, Vertiv (NYSE:VRT) is well-positioned to benefit. It sells the entire critical infrastructure for data centers and has 50 years of experience in the industry.
Vertiv’s advantages are clear: instead of offering one-point products, it provides a comprehensive end-to-end solution. First, it offers thermal management solutions that include liquid cooling technologies. Secondly, its portfolio includes data center power management products like transformers, distribution units, busways and microgrid integration capabilities. These solutions, plus its IT systems, offer an integrated end-to-end solution for CSPs.
After declining by over 15% over the past three months, Vertiv has become one of the interesting AI stocks to consider. After all, Q2 results highlighted the continued demand momentum across the business. Notably, this was another beat and raise quarter for Vertiv, as organic orders rose by 57% and organic net sales grew 14%.
In the quarter, adjusted operating margins hit 19.6%, a 510-basis points margin expansion. In response, the company raised its profit and free cash flow guidance for 2024. Now, management expects adjusted EPS between $2.47 – $2.53. Thus, using the low-end guidance, it trades at 31 times EPS, a fair multiple given the AI tailwinds ahead.
Nvidia (NVDA)
Today, most of the investment in AI is happening in the enabling layer – chip design, semiconductors and data centers. A significant portion is going towards building data centers powered by Nvidia chips. Nvidia is in a one-horse race because its Hopper and new Blackwell chips have a huge performance advantage.
Shifting the trillion-dollar installed base of traditional data centers to AI ones spells a multi-year tailwind for Nvidia. That’s why Bank of America doubled down on shares with a $150 price target. Analyst Vivek Arya notes that generative AI deployments are in the second year of a potential 3–5-year investment cycle.
Its new Blackwell chips are already in production and will be on the market later this year. This increases Nvidia’s revenue visibility into next year. Moreover, the company is combining its Grace CPU, Blackwell GPUs, NVLink, Quantum, Spectrum, mix and switches, high-speed interconnects and CUDA ecosystem to offer a complete solution to customers.
As of this writing, NVDA stock is one of the bargain AI stocks after pulling back from $140 to $113. At a forward non-GAAP price-to-earnings of 41, NVDA is trading below its 5-year average of 47. That’s too cheap for a stock expected to grow revenues by 97% this fiscal year.
Broadcom (AVGO)
In the AI chip race, Broadcom (NASDAQ:AVGO) has emerged as the runner-up behind Nvidia. The company is the leader in the custom AI chip market. For instance, it has long worked with Alphabet (NASDAQ:GOOG, NASDAQ:GOOGL) on its tensor processing units.
Due to the growing size and complexity of large language models, companies are opting for custom-designed AI chips to lower costs. Amazon (NASDAQ:AMZN) has launched its Tranium and Inferentia chips, while Microsoft (NASDAQ:MSFT) has its Maia chip.
Today, Broadcom counts Google as its largest customer, producing their TPUs. Besides, it’s already working with Microsoft and Meta Platforms (NASDAQ:META) on custom AI chips. Considering Broadcom is the leader in custom silicon, this is a massive tailwind.
At the same time, its AI networking revenues are soaring. In Q2 2024, the company saw strong demand from AI data center clusters for its switches, network interface cards and optical interconnects. The combined demand for AI networking and custom accelerators drove a 280% YOY surge in AI revenues to $3.1 billion.
Considering this acceleration in AI revenues, Broadcom is one of the best AI stocks. In FY2022, AI revenues were 10% of semiconductor revenues or about $2.5 billion. In FY2024, management expects this percentage to exceed 30% with AI sales of $11 billion, a quadruple from FY2022. This revenue momentum will continue into FY2025, making AVGO stock a buy.
On the date of publication, Charles Munyi did not hold (either directly or indirectly) any positions in the securities mentioned in this article. The opinions expressed in this article are those of the writer, subject to the InvestorPlace.com Publishing Guidelines.
On the date of publication, the responsible editor did not have (either directly or indirectly) any positions in the securities mentioned in this article.