To The Moon

Lip-Bu Tan’s move to lead Intel at 66 is more than a CEO change.

It is a high-risk turnaround bet on Intel’s future in:

• AI infrastructure

• CPU revival

• Advanced manufacturing

• Foundry services

• U.S. semiconductor strategy

1. Why He Took the Job

Tan originally planned to retire.

He had rejected Intel once in 2021 because of commitments at Cadence.

When Intel approached him again in 2025, many industry friends warned him not to take the role.

The risk was clear:

• Intel is extremely difficult to fix

• Failure could damage his reputation

• The company is strategically important to the U.S.

• The turnaround requires more than normal corporate restructuring

What changed his mind was a message from an old customer friend:

Intel is the foundation of the semiconductor industry. If you can save it, this is the major task to complete before retirement.

2. First Priority: Cash Flow 💰

Tan’s first focus was Intel’s balance sheet.

His logic:

Without healthy cash flow, Intel cannot fund advanced chips, fabs, packaging, or long-term R&D.

Key support mentioned:

• U.S. government became an important major shareholder

• Nvidia invested $5 billion

• That investment helped mobilize over $25 billion in broader resources

• SoftBank’s Masayoshi Son also provided support

Main point:

Tan is rebuilding Intel’s financial confidence first.

3. Internal Reform ✂️

Tan then moved to fix Intel’s organization.

His changes include:

• Cutting bureaucracy

• Making teams directly accountable

• Letting core engineering teams report directly to him

• Requiring negative customer feedback to be shared within 24 hours

• Replacing leaders if repeated tape-outs fail

• Restoring engineers’ importance inside Intel

• Embedding AI tools across R&D, production, finance, and operations

The message is simple:

Intel must become faster, more transparent, and more engineering-driven.

4. Stock Rebound 📈

Intel shareholders reportedly saw around a 6x market value return over the past 14 months.

But Tan said this is only the warm-up.

His long-term target:

10x shareholder return in 5 to 10 years.

He believes Intel cannot easily repeat Cadence’s 100x return because of its size, but a 10x roadmap is still achievable.

5. CPU Comeback 🔄

Tan’s most surprising view:

The CPU is making a comeback.

The AI market has focused heavily on GPUs, but Tan believes CPUs will matter more as AI shifts toward inference and agentic AI.

According to him:

• Training workloads are moving from CPU:GPU ratio of 1:8 toward 1:4

• Inference may move closer to 1:1

• Reinforcement learning and agent orchestration may favor CPUs

Why this matters:

Real-world AI needs more than training.

It also needs:

• Inference

• Scheduling

• Data calling

• Agent collaboration

• Edge deployment

• System coordination

That gives Intel a new opportunity.

6. Full-Stack AI Ambition 🧩

Intel does not only want to sell chips.

Tan wants Intel to build full-stack solutions, including:

• Silicon

• Software

• Systems

• Rack-level infrastructure

If Intel succeeds, its story could expand from “CPU company” to “AI infrastructure platform.”

7. Musk and Terafab 🚀

Tan also revealed Intel is working with Elon Musk on the Terafab project.

The logic:

AI demand is growing faster than semiconductor infrastructure.

Under the cooperation:

• Musk plans to build his own fab

• Intel provides technology and process support

• Tan meets Musk’s team every week

This could make Intel a deeper manufacturing partner for major AI players.

8. New Tech Bets 🧪

Intel is pushing its 18A and 14A roadmap while also planning future paths toward 1nm and 0.7nm.

But smaller nodes are becoming harder and more expensive.

So Intel is also investing in:

• EMIB advanced packaging

• Glass substrates

• GaN

• SiC

• InP

• Synthetic diamond

The future of chips may depend less on node shrinking alone and more on packaging, materials, power efficiency, and system design.

9. Foundry Trust 🤝

For Intel’s foundry business, Tan highlighted three key metrics:

• Yield

• Defect density

• Cycle time

These matter more than big promises.

Foundry is a trust business.

Customers will only give Intel major chip designs if Intel can prove stable quality, reliable delivery, and competitive timelines.

Tan expects Intel’s foundry potential to become clearer around 2030 to 2032.

10. Investor Watchlist 👀

Investors should watch five things:

• Can Intel repair its balance sheet?

• Can Intel rebuild engineering culture?

• Can CPUs regain value in AI inference?

• Can Intel win in packaging and new materials?

• Can Intel rebuild foundry trust?

Final Takeaway

Lip-Bu Tan taking over Intel at 66 is a bet on whether Intel can rebuild itself across finance, engineering, AI infrastructure, manufacturing, and foundry.

The upside is large, but the execution risk is equally high.

The real question:

Can Tan turn Intel’s comeback story into real results?

If yes, Intel may be entering a much larger long-term re-rating cycle.