Intel CEO gives investors a reality check

AI inventory buyers have primarily been skilled to observe a rule to put money into companies closest to GPUs, knowledge facilities, and hyperscaler spending.

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

Tan believes that the AI commerce is successfully transferring past a easy GPU story and into more durable infrastructure layers behind it

He successfully laid out a far larger map of the place the semiconductor area is being stretched.

That provides Intel a remarkably formidable investor narrative, but in addition a more durable check in proving it could possibly execute the place your entire AI provide chain is underneath strain.

Intel CEO factors to AI’s subsequent chip bottlenecks 

Intel CEO Lip-Bu Tan used the No Priors podcast to push the AI funding debate past GPUs.

Extra AI:

Maybe his largest rationale was that agentic AI is altering the compute combine.

Tan argued that the outdated coaching setup relied too closely on GPUs, however newer AI workloads require far more orchestration, reinforcement studying, and coordination throughout brokers. That’s bringing CPUs again into the dialogue. “Proper now, the demand may be very excessive for my CPU,” he stated.

For perspective, the AI commerce has been dominated by demand for accelerators.

Nonetheless, Tan feels the following section additionally is dependent upon the infrastructure across the accelerator: CPUs, reminiscence, interconnects, packaging, energy, and manufacturing capability.

He additionally flagged reminiscence as a strain level, saying “reminiscence is a much bigger scarcity” as corporations scramble for provide. Moreover, energy is one other constraint. Tan stated some international locations merely should not have sufficient energy capability to assist AI development, whereas energy conversion and thermal limits have gotten extra essential throughout the chip stack.

Furthermore, Tan stated a foundry is a “service enterprise” and a “belief enterprise”, the place prospects care about yield, defect density, cycle time, and reliability. If a chipmaker misses the mark on these factors, the client can lose gross sales.

Therefore, AI demand is way from being only a one-stock or one-chip story. Tan is pointing to a broader semiconductor bottleneck cycle, during which the winners are prone to be those who remedy the bodily limits constraining AI development.

Huge-bank Intel inventory targets present Wall Road is cut up after the AI rally

• Financial institution of America: $135. BofA double-upgraded Intel to Purchase from Underperform and lifted its goal from $96, citing stronger AI server CPU demand, foundry buyer momentum and better 2030 earnings energy.

• Citi: $130. Citi raised its Intel goal from $95, arguing that agentic AI may sharply broaden the server CPU market and provides Intel a much bigger data-center development runway.

• Mizuho: $135. Mizuho raised its goal to $128 whereas sustaining a Impartial score, citing a greater outlook for superior packaging, foundry platforms, and home chip manufacturing demand.

• Wells Fargo: $110. Wells Fargo lifted its goal from $85 however maintained an Equal Weight score, suggesting Intel’s AI CPU and foundry story is enhancing, however the inventory already displays a lot of that optimism.

• Barclays: $100. Barclays raised its goal from $65 whereas retaining an Equal Weight view, signaling extra confidence in Intel’s turnaround, however continued warning after the inventory’s big run.
  Sources: Yahoo Finance, TheStreet, Investopedia, Buying and selling View

Associated: Lip-Bu Tan’s web value: The Intel CEO’s wealth & revenue

What Intel nonetheless has to show in foundry and what comes subsequent for AI inventory buyers

Intel’s foundry story has clearly advanced over the previous 12 months or so.

The newest figures present why the market stays cautious.

In Q1 2026, Intel Foundry reported $5.4 billion of phase gross sales, up 16%, however intersegment eliminations have been $5.3 billion, which means the enterprise nonetheless largely helps Intel’s personal merchandise. The phase additionally posted a $2.4 billion working loss, whereas Intel Merchandise earned $4.1 billion.

The complete-year image is analogous.

Intel Foundry generated $17.8 billion of 2025 income, however solely $307 million got here from exterior prospects. Its working loss was $10.3 billion.

That’s the reason the following proof level is all about actual exterior buyer orders, produced at yields robust sufficient to assist quantity manufacturing.

There was some robust progress, although, of late.

Intel stated 18A-P entered threat manufacturing, with 9% higher efficiency on the identical energy, 18% decrease energy on the identical efficiency, and 20% to 40% higher thermal resistance.

Intel additionally introduced in Seok-Hee Lee to guide superior packaging and back-end manufacturing, areas that matter extra as AI chips change into multi-die methods.

For AI inventory buyers, the takeaway goes past Intel.

Tan’s message factors to a market transferring from GPU shortage to supply-chain shortage. CPUs, reminiscence, packaging, energy, and foundry capability have gotten investable bottlenecks. Naturally, Intel can profit from that shift, however provided that it proves prospects can belief its factories at scale.

Associated: Tesla rival says Elon Musk nonetheless has one main edge

This story was initially printed by TheStreet on Jun 22, 2026, the place it first appeared within the Investing part. Add TheStreet as a Most well-liked Supply by clicking right here.