Artifical Intelligence (AI) stocks have been the buzzword for investors who want to beat the S&P 500. Some AI stocks have produced tremendous returns in a short amount of time, but those same stocks can be quickly forgotten after a few bad weeks.
These three AI stocks have been relatively forgotten. They are small companies that have crushed the S&P 500 by wide margins, but these same growth stocks are down by more than 30% from their all-time highs. Yet, their fundamentals have still improved, making them due for rebounds.
Image source: Getty Images.
Penguin Solutions
Penguin Solutions (PENG +5.22%) is a memory-focused stock which develops valuable products that are used in AI infrastructure. More specifically, it's Compute Express Link (CXL) technology can improve AI inference performance while achieving lower costs than high-bandwidth memory (HBM) chips. The product stores memory from an HBM to ensure that the memory chip doesn't run out of storage so soon.
NASDAQ: PENG
Key Data Points
The company has some slow-growth legacy businesses that mask how much ground it is gaining in the memory industry. A 48% year-over-year revenue growth rate in its fiscal 2026 third quarter is already impressive, but investors should focus on the Integrated Memory segment when analyzing results. That part of the business more than doubled year over year and surged by 60% sequentially.
Penguin Solutions' other two segments also grew a little bit year over year but aren't worth focusing on. Notably, Integrated Memory, which includes CXL products, made up 57.5% of the company's total revenue. It was only 40.1% of revenue in the company's fiscal 2025 Q3. As this segment continues to soar, it will make up a larger percentage of total revenue and lift the company's overall sales.
MaxLinear
MaxLinear (MXL +7.54%) has more than quadrupled this year but remains down by more than 30% from its all-time high. A 55% year-over-year revenue jump in Q2 was a result of surging demand for MaxLinear's AI data center products.
CEO Kishore Seendripu cited the "strong momentum" in the company's optical AI data center business, which more than doubled year over year, for the successful results. Revenue also increased by 23% sequentially.
The company's guidance for Q3 also excited investors, with $215 million in revenue expected at the midpoint. That represents a 27% sequential improvement, showing that the AI build-out is meaningfully accelerating MaxLinear's business.
NASDAQ: MXL
Key Data Points
Seendripu specifically mentioned "the convergence of multiple growth drivers over the next two years," signaling that these high growth rates are here to stay. Investors were quick to accumulate shares when this news came out, but some investors have been taking profits, which may prove to be a mistake in the long run.
Aehr Test Systems
Aehr Test Systems (AEHR +0.94%) has been the best-performing growth stock of the bunch. It has almost produced a 5x return this year and is down by more than 30% from its all-time high.
At first glance, its revenue numbers look small compared to its market value. Aehr Test Systems made $50 million in its fiscal 2026, which was a slight drop from $59 million in its fiscal 2025. Meanwhile, the stock trades at a $3 billion market capitalization. However, the company anticipates $130 million to $150 million in fiscal 2027 revenue, which implies up to 3x year-over-year growth.
That's because the company specializes in test and burn-in solutions for semiconductor devices. Aehr runs these tests on chips to ensure they aren't defective when they reach customers' hands. The company has attracted hyperscalers and chipmakers alike since some tech giants are creating their own chips instead of exclusively relying on chipmakers.
NASDAQ: AEHR
Key Data Points
This materialized into a $22 million follow-on order from "a leading provider of advanced AI training and inference processors."
Aehr CEO Gayn Erickson cited the "complexity of advanced AI semiconductor packages" as a key catalyst for rising demand. The company is a silent play on the growing demand for semiconductors, and this broad term isn't limited to AI chips. It also includes central processing units (CPUs), memory chips, and network processors.
The stock certainly trades at a high valuation, but the projected revenue growth and long-term AI tailwinds have given investors good reasons for excitement.
