Intel CEO gives investors a reality check

AI stock investors are actually trained to follow the rule of thumb to invest in businesses around GPUs, data centers, and hyperscaler spending.

Intel ( INTC ) CEO Lip-Bu Tan recently flipped that script in the June 18 episode of No Priors: AI, Machine Learning, Tech, & Startups.

Tan believes that commercial AI success goes beyond a simple GPU issue and is a strong infrastructure layer behind it

He has successfully laid out a very large map where the semiconductor space is stretched.

That gives Intel a compelling investor narrative, but also a tough test to prove it can perform when the entire AI supply chain is under pressure.

Intel CEO identifies next AI bottlenecks

Intel CEO Lip-Bu Tan used the No Priors podcast to push the AI ​​investment debate beyond GPUs.

Additional AI:

Perhaps his biggest reason was that agent AI is changing the compute mix.

Tan argued that old training setups relied heavily on GPUs, but new AI workloads require more planning, reinforcement learning, and communication across agents. That’s bringing CPUs back into the conversation. “Right now, the demand is very high on my CPU,” he said.

In theory, the commercialization of AI is dominated by the demand for accelerators.

However, Tan feels that the next phase depends on the infrastructure around the accelerator: CPUs, memory, connectivity, packaging, power, and processing power.

He also flagged memory as a pressing issue, saying “memory is in short supply” as companies scramble for supplies. Additionally, energy is another constraint. Tan said that some countries do not have enough power to support the growth of AI, while power conversion and thermal limitations are becoming more important for the entire chip stack.

In addition, Tan said the center is a “service business” and a “trust business”, where customers care about yield, error density, cycle time, and reliability. If the chip maker misses the mark on those points, the customer could lose the sale.

Therefore, the demand for AI is far from being just a matter of a single stock or a single chip. Tan points to a broader semiconductor bottleneck cycle, where the winners are likely to be those who solve the physical limitations that hinder the growth of AI.

Big bank Intel’s stock target shows Wall Street is divided after the AI ​​rally

• Bank of AmericaPrice: $135. BofA upgraded Intel twice to Buy from Underperform and raised its target from $96, citing strong demand for AI server CPUs, momentum from existing customers and higher earnings potential for 2030.

• Citi: $130. Citi raised its Intel target from $95, arguing that agent AI could significantly expand the server CPU market and give Intel a major growth channel for the data center.

• MizuhoPrice: $135. Mizuho raised his target to $128 while maintaining a neutral rating, citing a positive outlook for advanced packaging, core platforms, and domestic chip manufacturing demand.

• Wells Fargo: $110. Wells Fargo raised its target from $85 but maintained an Equal Weight rating, suggesting Intel’s AI CPU and Foundry story is improving, but the stock is already showing much of that optimism.

• Barclays: $100. Barclays raised its target from $65 while maintaining an Equal Weight outlook, indicating more confidence in Intel’s turnaround, but remained cautious after the stock’s big run.
  Sources: Yahoo Finance, TheStreet, Investopedia, Trading View

Related: Lip-Bu Tan net worth: Intel CEO net worth and income

What Intel still has to prove in its innovation and what’s next for AI stock investors

Intel’s innovation story has clearly evolved over the past year or so.

The latest statistics show why the market remains cautious.

In Q1 2026, Intel Foundry reported $5.4 billion in segment sales, up 16%, but segment completion was $5.3 billion, which means the business still heavily supports Intel products. The segment also posted an operating loss of $2.4 billion, while Intel Products earned $4.1 billion.

The picture for the full year is the same.

Intel Foundry has generated $17.8 billion in revenue by 2025, but only $307 million comes from foreign customers. Its operating loss was $10.3 billion.

That’s why the next proof point is about actual orders from foreign customers, which are produced at a yield strong enough to support volume production.

There has been some solid progress, however, recently.

Intel said the 18A-P went into production risk, with 9% better performance at the same power, 18% lower power at the same power, and 20% to 40% better heat resistance.

Intel also brought on Seok-Hee Lee to lead advanced integration and back-end manufacturing, areas that are increasingly important as AI chips become multi-die systems.

For AI stock investors, the pick goes beyond Intel.

Tan’s message points to the market moving from GPU shortages to supply chain shortages. CPUs, memory, packaging, power, and Foundry capabilities become the constraints that can be deployed. Naturally, Intel could benefit from that change, but only if it proves that customers can trust its factories at scale.

Related: Tesla’s rival says Elon Musk still has one big advantage

This story was originally published by TheStreet on Jun 22, 2026, where it appeared first in the investing category. Add TheStreet as a favorite source by clicking here.