2 AI stocks split into stocks to buy before they rise 165% and 245%, according to certain Wall Street analysts

Year to date, Nvidia (NVDA 1.67%) and Super Micro Computer (SMCI 1.70%) were the best performing stocks of Nasdaq 100 and two of the three best performing stocks in the S&P 500. Both companies chose to split their shares to make the stock more affordable. Specifically, Nvidia completed a 10-for-1 stock split in June, and Supermicro has a 10-for-1 stock split planned for September.

Somewhat surprisingly, Wall Street remains bullish on both companies. The average price target implies a 21% upside for Nvidia and a 22% upside for Supermicro, but some analysts see much bigger gains on the horizon.

• In June, Beth Kindig of I/O Fund published an analysis in Forbes which values ​​Nvidia at $10 trillion by 2030, echoing a prediction Jim Cramer made two years ago. That’s a 245% upside to its current market cap of $2.9 trillion.

• In April, Loop Capital’s Ananda Baruah raised his 12-month price target for Supermicro to $1,500 per share, implying a 165% upside from the current price of $567 per share. Similarly, Rosenblatt’s Hans Mosesmann set his price target at $1,300 per share, implying a 129% upside.

Here’s what investors should know about these artificial intelligence (AI) stocks.

1. Nvidia

Nvidia graphics processing units (GPUs) were originally designed to render stunning computer graphics for video games and 3D design applications. But the company repurposed its GPUs as data center accelerators when it released its CUDA parallel computing platform in 2006. CUDA has evolved into a robust ecosystem of software tools that streamline the development of GPU-accelerated applications across disciplines, from computational chemistry to artificial intelligence. .

Nvidia dominates the data center accelerator market. The company accounted for 98 percent of data center GPU shipments in 2023, according to semiconductor analysts TechInsights. Nvidia also has over 80% market share in AI processors.

This dominance is partly due to superior performance. Nvidia GPUs consistently outperform competing chips in MLPerf benchmarks, tests that provide unbiased evaluations of AI systems through training and inference.

However, the company is truly formidable because it offers a full-stack computing solution — meaning it combines the hardware, software, and services that businesses need to build, deploy, and manage AI applications. I don’t just mean GPU and CUDA. Nvidia also provides additional data center hardware such as networking equipment and central processing units (CPUs) and offers a comprehensive AI-as-a-service solution called DGX Cloud.

Nvidia reported better-than-expected financial results in the first quarter of fiscal 2025 (ended April 2024). Revenue rose 262% to $26 billion on a strong drive from the data center segment, and non-GAAP net income rose 461% to $6.12 per diluted share. CEO Jensen Huang said, “Our data center growth has been fueled by strong and accelerating demand for generative AI training and inference on the Hopper platform.”

Wall Street expects Nvidia to grow adjusted earnings 52% annually through fiscal 2026 (ending January 2026). That consensus estimate makes its current valuation of 57.7 times adjusted earnings look reasonable.

Investors interested in buying Nvidia stock should start with a small position today. If the stock falls following the upcoming earnings report on August 28th, consider using this opportunity to build a slightly larger position.

Ultimately, I think Nvidia could be a $10 trillion company, but I’m skeptical of reaching that milestone by 2030.

2. Super Micro Computer

Supermicro manufactures high-performance computing platforms, including storage solutions and servers optimized for intensive workloads such as data analytics and artificial intelligence. The company holds a leading position in the AI ​​server market thanks to its modular approach to product design and its in-house manufacturing capabilities.

Specifically, Supermicro makes electronic “building blocks” that can be assembled into servers in an almost endless number of combinations. Rivals offer customers a more limited menu,” The Wall Street Journal. The company also does most of its research, development and assembly at its Silicon Valley facilities, which support the rapid release of servers with the latest from vendors such as Nvidia.

Supermicro reported mixed financial results in the fourth quarter of fiscal 2024 (ended June 30). Revenue rose 143% to $5.3 billion on record demand for AI infrastructure. But non-GAAP net income rose just 78% to $6.25 per diluted share as costs associated with direct liquid cooling (DLC) components pressured margins.

Wall Street had anticipated adjusted earnings growth of 130%. That mistake sent the stock down 17% following the report.

However, management provided important context for the earnings call. While gross profit margin fell 5.8 percentage points to 11.2% in the fourth quarter, Chief Financial Officer David Weigand told analysts that figure should normalize between 14% and 17% by end of fiscal year 2025 as DLC production capacity ramps up. Furthermore, investing in DLC could help Supermicro gain share in AI servers.

Liquid-cooled AI servers reduce data center energy consumption. So demand for DLC solutions is expected to grow rapidly, accounting for at least 15% of all data center installations in the next two years, up from less than 1% in the past. Supermicro has emerged as an early leader in DLC solutions, which could ultimately drive demand for its AI servers.

Wall Street expects Supermicro to grow adjusted earnings by 41% annually through fiscal 2026. That makes the current valuation of 25.7 times adjusted earnings look cheap. That said, shares could fall further if Supermicro misses Wall Street’s earnings estimates in the coming quarters.

Investors who are comfortable with that risk could buy a small position today, but not with the expectation of triple-digit gains over the next 12 months. It could happen, but anyone banking on that result is begging to be disappointed.