Intel has become a star of the artificial intelligence trade, and Jim Cramer said investors need to own the stock.

“If you don’t own Intel, please buy it,” said Jim during Thursday’s Morning Meeting, touting the chipmaker as it becomes more crucial to the artificial intelligence data center buildout. He said it’s now his favorite name in the chip space.

“Intel’s my No. 1 name, not Nvidia ,” Jim said. He argued that portfolio holding Nvidia, the leading maker of AI chips, has become a source of funds for investors looking to raise capital ahead of the SpaceX initial public offering Friday. But with Intel, that’s not the case. 

Intel’s stock popped up nearly 8% Thursday following a double upgrade by Bank of America. The firm elevated the stock to a buy from a sell-equivalent rating, and raised its price target to $135 from $96. Analysts based the call on higher confidence in Intel’s opportunity to benefit from AI spending in two ways: its central processing units (CPUs) and its contract chip manufacturing business, known as Intel Foundry.

Both reasons are why the Club initiateda position in the company last week at a price target of $140 per share. Shares of Intel and other AI chipmakers were then hit with a wave of selling in recent days. Thursday’s gains put Intel back to around $115, slightly above the price of our first buy on June 3. We added to our position at $101.80 a share on June 5.

Bank of America now projects Intel’s server CPU sales to hit $40 billion by calendar year 2030 — a very bullish target versus the consensus. Wall Street currently expects Intel’s data center and AI segment to generate $32.5 billion in 2030, according to FactSet; the vast majority of that segment’s revenues are from CPUs, per BofA. The analysts believe Intel will capture about 25% of a total addressable market for server CPUs worth $170 billion, up from its prior estimate of $135 billion.

Intel’s blowout quarterin April was a huge reality check for investors that revealed a booming demand for CPUs. Nvidia’s graphic processing units (GPUs) have long dominated artificial intelligence computing and remain key for training AI models. But as more AI computing moves from training to the daily inference stage — the process of AI models answering queries and reacting to user demands — CPUs are becoming hot commodities for data centers. The growth of agentic AI systems, in particular, which are capable of autonomously completing tasks for users, require a lot of CPUs. We bought Arm Holdings as a beneficiary of this shift too. Bank of America also raised its price target on Arm to $335 from $245.

Intel’s chip manufacturing business is increasingly positioned to help relieve supply chain bottlenecks. Taiwan Semiconductor Manufacturing Co. , the largest contract chip manufacturer, is essentially at full capacity and expanding its footprint takes time. BofA sees growing opportunities for Intel Foundry to step up as a TSMC alternative.