Good morning everyone.

Today, I will walk you through the financial valuation of Intel and the
rationale behind our key assumptions. Our analysis is focused on determining whether NVIDIA
should acquire Intel, and based on our valuation, we conclude that Intel is overvalued, making
the deal financially unattractive. Let's break down the financial valuation and key assumptions
step by step.

Intel’s revenue is projected to grow from $55.3 billion in 2024 to $67.9 billion in 2028,
reflecting a compound annual growth rate (CAGR) of ~5.2%. This assumption is based on Intel’s
2022 investor meeting, where management targeted an acceleration to 10%-12% year-over-year
(YoY) growth by 2026. However, we have taken a more conservative approach, considering
Intel’s declining market share and competitive pressures from AMD and NVIDIA in AI chips.
AMD now holds 25% of the CPU market, eroding Intel’s dominance, while NVIDIA controls
80% of the AI GPU market, leaving Intel with minimal foothold in high-growth AI segments.
While Intel aims for aggressive revenue growth, our model reflects the challenges in execution.

The decline in COGS as a percentage of revenue—from 60% in 2023 to 46% by 2028—is driven
by operational efficiency and product mix improvements. Intel’s $50 billion CHIPS Act funding
allows it to bring 20-30% of production in-house by 2026, reducing outsourcing costs to TSMC.
Additionally, higher-margin segments foundry services will help in bringing the cost down.

For SG&A and Research & Development (R&D) expenses, we have assumed them as a
percentage of revenue based on historical trends. The rationale behind this is that Intel has
maintained relatively stable SG&A and R&D expenditure patterns as a proportion of revenue
over time. This consistency ensures that our model captures the inherent cost structure of the
business without making speculative reductions or increases.

Our model projects net income to grow significantly from $2.1 billion in 2024 to $8.9 billion in
2028, representing a CAGR of ~40%. This sharp increase is primarily driven by expected
improvements in operating efficiency and cost reductions. However, this assumes that Intel
successfully executes its strategic initiatives without major setbacks in its foundry business.
However, this assumes no delays in fab upgrades or product launches—a risky bet given Intel’s
history of missed deadlines

The WACC of 9.59% (from the appendix) incorporates Intel’s debt-heavy capital structure and
equity risk. The terminal growth rate of 3.25% which aligns with semiconductor industry norms
(2-4%) but avoiding overly optimistic “high-growth” assumptions.

Our valuation model projects Intel’s Free Cash Flow (FCF) to decline from $7.5 billion in FY25
to $6.5 billion in FY27, reflecting operational challenges and capital expenditure requirements.
This results in an enterprise value of $90.2 billion and a terminal value of $68.1 billion. After
adjusting for net debt, the final equity value stands at $48.7 billion, translating to a fair price per
share of $11.30. This represents a downside of 45.7% from Intel’s current valuation, highlighting
that Intel is overvalued and the acquisition is costly for NVIDIA.
Q1: Your model assumes a 5.2% revenue CAGR for Intel, far below its 10-12% target.
Why not factor in management’s optimism?

Answer:
Intel’s 10-12% target assumes flawless execution, but our conservatism reflects three realities:

1. Market Share Loss: AMD now holds 25% of the CPU market (Slide 5), and NVIDIA
dominates 80% of AI GPUs, leaving Intel with limited growth avenues.
2. Foundry Delays: Intel’s foundry revenue is just 15% of total sales (Slide 7), and
catching up to TSMC’s 3nm/5nm nodes requires $50B+ in capex over 5+ years (Slide
15).
3. AI Weakness: Intel’s AI accelerators hold <5% market share (Slide 5), with no near-term
path to challenge NVIDIA or AMD.
We prioritized execution risks over aspirational targets.

Q2: COGS as a % of revenue drops sharply to 46%. Is this realistic given Intel’s reliance
on TSMC?

Answer:
Yes, for two reasons:

1. CHIPS Act Impact: Intel’s $50B+ in subsidies (Slide 7) funds in-house production—
reducing TSMC outsourcing costs by 20-30% by 2026 (Slide 15).
2. Product Mix Shift: Higher-margin AI/foundry services (15% YoY growth) dilute the
low-margin CPU segment (Slide 7). Even Intel’s 2022 roadmap emphasizes this shift.

Q3: Why keep R&D at 30% of revenue? Shouldn’t Intel cut R&D to improve margins?

Answer:
Cutting R&D would be catastrophic:

 Technological Lag: Intel already lags TSMC in advanced nodes (Slide 7). Reducing
R&D would widen the gap, making foundry modernization impossible.
 Competitive Necessity: NVIDIA spends 20% of revenue on R&D (Slide 6). Intel’s 30%
reflects its catch-up race in AI and nodes—a survival cost, not discretionary spend.

Q4: Terminal value accounts for 75% of Intel’s valuation. Isn’t this overly optimistic?

Answer:
The high terminal value reflects two deliberate choices:
 1. Conservative Growth: We used a 3.25% terminal rate (Slide 24), aligning with
semiconductor industry norms (2-4%) and avoiding aggressive assumptions.
2. Capex Normalization: Post-2028, Intel’s fab upgrades are complete, freeing cash flow.
However, this still implies a 5-7 year turnaround—time NVIDIA cannot afford (Slide
19).

Q5: Why not acquire Intel for its fabs? NVIDIA could vertically integrate like Apple-Intel.

Answer:
Three reasons:

1. Outdated Fabs: Intel’s fabs are 2-3 years behind TSMC (Slide 7). Modernizing them
requires $20B+ annual capex (Slide 9), straining NVIDIA’s balance sheet.
2. Regulatory Risk: The FTC/EU would likely block this deal due to monopoly concerns in
GPUs + CPUs (Slide 13).
3. Better Alternatives: Partnering with TSMC (Slide 18) secures advanced nodes without
$90B in debt or dilution.

Q6: Intel’s net income grows 40% CAGR. Isn’t this contradictory to declining FCF?

Answer:
Not at all. Net income growth is driven by margin expansion (COGS decline), but FCF declines
due to:

 Heavy Capex: $20B+ annually to modernize fabs (Slide 9).

Q7: What if Intel’s AI accelerators gain traction? Would that change your valuation?

Answer:
Unlikely, given two structural issues:

1. Market Position: NVIDIA’s CUDA ecosystem and 80% AI GPU share (Slide 5) create
insurmountable barriers.
2. R&D Mismatch: Intel spends 30% of revenue on R&D but prioritizes CPUs/foundries
(Slide 7)—not AI. Even if AI revenue doubles, it would only
add 2−3Bannually,immaterialagainst2−3Bannually,immaterialagainst90B acquisition
costs.
Q8: Why not use a lower WACC to boost Intel’s valuation?

Answer:
A 9.59% WACC (Slide 24) already balances:

 Debt Risk: Intel’s D/E ratio of 0.55 and 4-7% cost of debt.
 Equity Risk: Declining market share justifies a 5.6% equity risk premium.
Using NVIDIA’s WACC (8%) would ignore Intel’s standalone risks.

Q9: How does the US-China chip war impact your model?

Answer:
It reinforces our thesis:

 Supply Chain Risk: Intel relies on TSMC for advanced nodes (Slide 6), which could
face China-Taiwan disruptions.
 Opportunity Cost: The CHIPS Act funds TSMC/Samsung fabs in the US too (Slide 4),
reducing Intel’s unique advantage.