2 Top Artificial Intelligence (AI) Stocks Ready for a Bull Run

The bull market continues to roar higher, led by artificial intelligence (AI) stocks. As AI infrastructure spending continues to soar, bottlenecks in the industry are starting to develop, and that's creating clear-cut opportunities for some companies to benefit.

Let's look at two AI stocks ready for bull runs due to these bottlenecks.

Image source: Getty Images.

1. Micron Technology

One of the largest AI bottlenecks right now is memory, and Micron Technology (NASDAQ: MU) is one of the best-positioned companies to benefit from the current dynamics in the memory industry. Graphics processing units (GPUs) and other AI chips depend on high-bandwidth memory (HBM) to operate to their fullest potential.

As such, HBM modules are essential for high-bandwidth activities like large language model (LLM) training and AI inference. Given the boom in data center infrastructure spending, it should be no surprise that HBM demand is currently through the roof.

However, this is also helping lead to supply shortages across the entire memory chain, which is driving up both DRAM (dynamic random access memory) and NAND (flash memory) prices. DRAM is used for short-term memory due to its speed, and HBM is a specialized form of DRAM that has even higher bandwidth and is more energy efficient.

Given the need for HBM, memory suppliers are focused on this market; however, HBM needs up to four times the wafer capacity of regular DRAM. This is leading to lower overall DRAM supply and increased prices.

At the same time, NAND, which is used for long-term storage, is also in short supply. NAND prices crashed a few years ago, and memory companies have been reluctant to bring back production. However, AI servers need massive, high-performance solid-state drives (SSDs) using flash memory, because HBM alone can't support the massive amount of data being processed.

Micron is one of the best ways to play the current dynamics in the memory market, because it is involved in both the DRAM and NAND markets. The company's entire HBM capacity is sold out for this year, and it has doubled its construction capital expenditures (capex) and is working to increase production for both HBM and DRAM. Its NAND production is oversubscribed as well. Micron is already seeing the benefits of increased prices, and that should continue to be a huge driver for the company for the foreseeable future.

2. Applied Digital

Another big bottleneck in AI infrastructure is access to cheap power. That dynamic is benefiting Applied Digital (NASDAQ: APLD), which builds and operates data centers specifically designed for training LLMs and inference.

The company was previously a Bitcoin miner, which is where it learned the ins and outs of sourcing cheap power and building out large hosting campuses. It has now turned its focus to AI data centers. After spinning off its cloud computing business and merging it with EKSO Bionics Holding to form a new company called ChronoScale, Applied Digital is looking to convert to a data center real estate investment trust (REIT).

Applied Digital has secured financing and is aggressively building out new facilities. It just completed the first building, supporting 100 megawatts of power, at its Polaris Forge 1 campus, which is being built to support CoreWeave. That campus will grow to 400 megawatts by 2027, and it has also added a 200-megawatt customer at its Polaris Forge 2 campus. Earlier this month, the company noted that it was in advanced talks with a couple of large cloud computing providers to provide them with 900 megawatts of power across two to three sites.

Applied Digital has huge growth ahead of it, and the company looks well-positioned, given its backing from Macquarie Group, which is helping provide much of the financing for these large data center projects.

Geoffrey Seiler has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Bitcoin and Macquarie Group. The Motley Fool recommends Micron Technology. The Motley Fool has a disclosure policy.