2 AI Stocks Jim Cramer Says Are Still Worth Buying After Massive Gains

The former hedge fund manager Jim Cramer has been making bold stock calls on business-focused cable channel CNBC since the 1990s. In the mid-2000s, he launched his now-flagship show Mad Money, which airs weeknights at 6 p.m. ET.

Cramer likes to go where the action is, so it's only natural that he's been talking a lot about artificial intelligence (AI) stocks, which have simply stolen the show over the past few years and essentially become the market, for better or worse.

AI stocks can be difficult to assess because while many are extraordinarily promising companies, many have also made big gains, which could tempt investors to lock in gains, especially with lots of market uncertainty right now.

Here are two AI stocks that Cramer still likes after massive gains.

Image source: Getty Images.

Nvidia -- 1,030%-plus gain in five years

Few stocks have been bigger winners than Nvidia (NASDAQ: NVDA), which is now the largest publicly traded company by market cap at $5.57 trillion.

Nvidia's general-purpose graphics processing units (GPUs), which are vital to training AI models, have dominated this market, with the company boasting gross margins in the mid-70% range. Another key advantage for the company is its CUDA software layer, which Nvidia has built a strong network around over the past two decades.

CUDA is closely integrated with Nvidia GPUs, which makes disrupting the company's market position more difficult than it appears.

Cramer has long been bullish on Nvidia, recently calling it "the best investment in the world" and recommending investors "Own it. Don't trade it." And this is after Nvidia has generated over a 1,030% return over the past five years, and a nearly 24% gain this year, which is big for a stock at its elevated market cap.

Cramer has recently been pushing Nvidia to implement a half-trillion-share repurchase plan.

While they didn't go that far, they recently approved the largest increase in history to their share repurchase authorization, raising Nvidia's remaining authorization by $150 billion to $235 billion, which they expect to use between now and the end of their fiscal year 2028.

"But this is a very significant buyback, and I think if they're active and in there every day, it will change the trajectory of the stock," Cramer said recently. Nvidia doesn't look expensive on the surface. It trades at less than 25 times forward earnings and is guiding for 70% revenue growth in its next fiscal year.

But one thing that is crucial for Nvidia -- and that investors should be constantly monitoring -- is how long the intense AI spending can continue. Nvidia needs companies to keep building data centers to house its chips, so as long as this spending continues, Nvidia will keep winning.

Micron -- 1,410%-plus gain in five years

Another AI stock Cramer is sticking with is the memory maker Micron Technology (NASDAQ: MU), which is up over 1,410% in the past five years and roughly 485% over the past year. Memory has arguably been the biggest AI winner of the year because it's so supply-constrained.

Micron makes NAND flash memory and dynamic random-access memory (DRAM), both of which play crucial roles in supplying GPUs with data that enables AI models to respond to queries so quickly and with such complexity.

NAND is for storing large-scale data sets, while DRAM is the foundation for high bandwidth memory that models need to quickly answer queries. As AI models scale and become more complex, GPUs need more memory, which is why they are in such high demand.

"While I acknowledge that I am not early, I do not think I am late," Cramer said in August, regarding memory stocks. "I think Micron can double again before the boom comes to an end, assuming there's no data center slowdown."

Cramer also said he likes that the memory makers have been returning capital to shareholders, rather than just investing in new capacity.

In a mid-September Mad Money broadcast, a caller asked Cramer whether to take gains on the stock, and he said he still thought it was cheap based on forward earnings.

Micron recently reported earnings that once again blew past analyst estimates, and it also provided guidance for the current quarter, which came in higher than expected.

While the stock's reaction was muted, Cramer reiterated his call to hold Micron. Although Cramer is right that Micron trades cheaply at just 6.7 times forward earnings, memory makers usually trade at low multiples because they are viewed as cyclical.

Typically, by the time memory makers catch up with demand, demand softens, and then they end up with a supply glut. A cohort of investors thinks the AI supercycle may be different.

I think the memory makers are a bit harder to call when it comes to figuring out when the companies will no longer be supply-constrained. But the stock should continue to perform well as long as the AI trade holds up.

Overall, I'm more cautious on this one after the massive run already. If you do invest, please definitely practice dollar-cost averaging for this one.

Bram Berkowitz has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Micron Technology and Nvidia. The Motley Fool has a disclosure policy.