On Mad Money, a recent caller pressed Jim Cramer on Cerebras Systems (NASDAQ:CBRS) with a pointed setup: “Last week, CrowdStrike chose Cerebras to power their real-time Falcon AI detection. When George Kurtz vouches for your inference feed, isn’t it time to stop treating Cerebras like a post-IPO trade?”
Cramer conceded the valuation math. “It’s certainly reasonable to say it’s down so much and the P/E multiple is not that high,” he said. Then he redirected the money. “I don’t want to buy a lot of tech. The only one that I’m currently buying is Intel, which I think has better prospects than Cerebras. But I like your logic.”
That is a narrow, deliberate call. Cramer is holding his tech book steady and directing fresh capital toward the legacy incumbent, which he sees as the one AI semiconductor name earning it right now over the hyper-growth pure play.
Why Intel Is Cramer’s Pick
Intel (NASDAQ:INTC | INTC Price Prediction) delivered the kind of quarter that supports Cramer’s conviction. Q2 FY2026 revenue landed at $16.13 billion, up 25.4% year over year, described by CEO Lip-Bu Tan as “strongest revenue growth in more than 15 years.” Non-GAAP EPS of 42 cents beat the 21-cent estimate by 93.1%, a swing from prior-year losses to real operating profit.
The AI story is now on Intel’s income statement. The Data Center and AI segment posted $6.26 billion in revenue, up 59% year over year, and Intel Foundry grew 31% even as it absorbed a $2.1 billion quarterly operating loss. Tan framed the setup this way: “AI is driving unprecedented demand for compute, and as we continue to execute, Intel is well-positioned to capture sustainable growth across our CPU franchise, ASICs, advanced packaging and vast wafer foundry network.” The full release is available in Intel’s Q2 2026 8-K filing.
The stock has already run hard. INTC is up more than 118% year to date and 321% over the past year, trading around $85.97 on July 29. The stock is down 35% over the past month, though, and its forward P/E sits at 105, with the average analyst target at $115.65. Backing the turnaround: an NVIDIA (NASDAQ:NVDA) $5 billion equity investment, SoftBank’s $2 billion stake and a U.S. government equity position tied to CHIPS Act funding.
The Cerebras Case Cramer Won’t Take
Cerebras is the harder trade to justify on fundamentals despite genuine momentum. Q1 FY2026 revenue was $193.4 million, up 94% year over year, driven by a Cloud and Other Services segment that grew 178%. Yet the company posted a GAAP loss of 22 cents per share and carries negative shareholders’ equity of -$194.7 million.
Guidance is where the margin story gets uncomfortable. Management pointed Q2 core gross margin to 36-38%, with core operating margins of -30% to -32%. Full-year 2026 core revenue is guided to $855 million to $865 million. CEO Andrew Feldman’s framing remains aggressive: “Cerebras’ wafer-scale technology delivers the fastest AI in the world. And fast AI is more valuable than slow AI because it is more productive.” Details are in the company’s Q1 2026 8-K.
Cramer’s comment about the multiple has data behind it. CBRS traded around $180.69 on July 29, down 35.20% from its post-IPO peak of $311.07 on May 14. The $20 billion, 750MW OpenAI inference deal is real, and so is customer concentration risk. That is the fork Cramer is trading around.
What Investors Should Watch
Cramer is threading a specific needle: buying the profitable-turnaround AI beneficiary with government and NVIDIA backing while acknowledging Cerebras’s valuation reset. Reddit sentiment on INTC has skewed bearish to very bearish over the past 30 days, with retail investors flagging dot-com era valuations. That makes his call a contrarian one, anchored to Intel’s Q3 guide of $15.8 billion to $16.8 billion in revenue and non-GAAP EPS of 38 cents. The next earnings report will tell investors whether the DCAI curve keeps bending up or whether Cerebras’s speed-of-inference pitch starts pulling incumbent CPU dollars.
Contact [email protected] for any questions or corrections.