There’s a version of this story that writes itself. Intel, once left for dead, claws back from the edge. A new CEO. A government backstop. NVIDIA calling with a $5 billion check. The stock up 85% year to date. Headline writers love it. The problem is that markets don’t reward good stories. They reward data that surprises relative to expectation. And right now, expectation for INTC is running dangerously hot.

Earnings drop Thursday, April 23, after market close. Consensus sits around $0.01 non-GAAP EPS for Q1 2026. The stock was trading near $68 heading into the print — a price that implies the market has already decided the turnaround is real. Those two things are in tension. Active traders need to understand exactly where that tension lives before Thursday’s number hits.

The Numbers, Without the Spin

Intel closed fiscal 2025 with full-year revenue of $52.9 billion, flat year over year. Not contracting, but not growing either. For a company burning through billions in foundry investment, flat top-line is a holding pattern, not a recovery. The segment breakdown adds nuance. Data Center and AI posted Q4 revenue of $4.7 billion, up 8.9% year over year — its strongest quarterly growth in a decade. The Client Computing Group declined 7% in the quarter. The PC business is still dragging. And Intel Foundry posted an operating loss of $2.51 billion in Q4 alone.

Now look at the Q4 beat, because the options market is referencing it. Intel reported Q4 EPS of $0.15 against analyst expectations of $0.04 — a 275% beat. The stock sold off 22% over the following two sessions. The guidance is what did it. Intel guided Q1 revenue of $11.7 billion to $12.7 billion with non-GAAP EPS of approximately $0. The market punished the print before recovering in the weeks that followed as new catalysts emerged. That January lesson is the most relevant framework for how to think about Thursday.

For Q1, consensus sits at roughly $12.32 billion in revenue. The gross margin guide of approximately 34.5% — down from 37.9% in Q4 — is the sleeper number. Intel’s internal target is 40% gross margin before setting a higher threshold. Until that level is reached, every quarter is a story of potential rather than proof. The company ended 2025 with $14.27 billion in cash, up 73% year over year, and cut headcount from 108,900 to 85,100. The restructuring is real. The question is whether it converts to margin fast enough.

The AI Spending Question

Is Intel’s AI spending out of control? The honest answer is: not out of control, but not yet productive at the scale the investment requires. Intel fell short of its goal to sell $500 million worth of AI accelerators in 2024. The Gaudi software ecosystem remains the core problem — developers who have built workflows on NVIDIA’s CUDA stack for years don’t have native familiarity with Gaudi’s architecture. That gap doesn’t close overnight. Gaudi 4 has captured roughly 6% of the data center accelerator market, positioning itself as a value-per-dollar alternative. That’s real. It’s also not transformative.

What matters is the 22% figure — Intel’s share of broader data center AI when CPUs are included alongside accelerators. The real AI opportunity for Intel was never about displacing NVIDIA in GPUs. It was always about Xeon in the same rack. Which is exactly what the NVIDIA partnership is structured around.

18A and the NVIDIA Deal: What’s Real

The 18A node is the fulcrum of the entire Intel thesis. Yields have improved — running around 7% monthly improvement over the past several months, which puts Intel on the industry-standard ramp curve. That’s meaningful. It means the early chaos has been replaced by something that resembles a predictable manufacturing process. But Intel’s CFO has been explicit: yields are sufficient to support Panther Lake shipments, not yet high enough to deliver normal profit margins. The desired cost level arrives by end of 2026. Industry-standard results in 2027. This is a 2027 story dressed up in 2026 headlines.

The NVIDIA deal is real and it matters — but not in the way the stock price suggests. NVIDIA purchased $5 billion in Intel equity and agreed to co-develop CPUs and SoCs that integrate NVIDIA RTX chiplets with Intel x86 architecture. The partnership creates native integration between NVIDIA GPUs and Intel CPUs in AI infrastructure stacks. That’s legitimate. What it is not: a wafer commitment. NVIDIA evaluated the 18A node and did not proceed to mass production. The deal gives Intel strategic relevance in NVIDIA-led infrastructure. It does not fill the foundry’s capacity utilization gap. The first co-developed products are expected in late 2026 to early 2027. That’s the timeline to hold management against on Thursday’s call.

Options Market: What the Market Is Pricing

Intel’s 30-day implied volatility heading into the week of April 16 was at 76 — near the top of its 52-week range of 38 to 77. Options are not cheap. For a stock near $68, IV at that level implies an expected move of roughly 10 to 12% in either direction around the event. The call-to-put ratio of 1.9:1 is modestly skewed toward calls, consistent with a stock that has already run hard. This doesn’t look like aggressive directional conviction — it looks like hedged upside positioning.

At IV near 52-week highs, buying outright calls or puts into Thursday means paying a significant premium for a move that may or may not materialize. The January print — 275% EPS beat, 22% selloff — is the reference point. The direction surprised most buyers. With IV rank near the top of the annual range, the environment favors premium collection over premium payment. Defined-risk structures that benefit from volatility compression post-event are the more relevant framework here than naked directional plays.

If you believe Intel beats on revenue and guidance: A bull call spread — buying the April 25 $70 call, selling the $77.50 call — captures upside while limiting cost into elevated IV. Max profit above $77.50. Max loss is the net premium paid.

If you believe Intel misses on guidance or margin: A bear put spread — buying the April 25 $65 put, selling the $57.50 put — defines risk at the premium paid and targets a return toward pre-NVIDIA-deal price levels.

For traders expecting a large move but uncertain on direction: A short iron condor — selling the $77.50 call and $57.50 put, buying the $82.50 call and $52.50 put as wings — collects net premium if INTC stays within a defined range. This is the structure that benefits most directly from IV crush. Post-earnings IV compression typically plays out within the first 30 to 60 minutes of the next session. Plan the exit before the event, not after.

What to Watch Thursday

• Q2 guidance vs. the $12.32B consensus: The Q1 revenue softness was pre-telegraphed. The market is trading the forward guide, not the reported quarter.
• Gross margin trajectory: Any language accelerating the path toward 40% is a bullish signal. Continuation of the 34.5% narrative without a clearer timeline is not.
• 18A yield language: “Pipeline” vs. “commitments” from external customers is a meaningful distinction. Listen for specifics on Microsoft and Amazon 18A programs.
• NVIDIA milestone updates: Specific product delivery dates for the first CPU-GPU SoC carry more weight than general partnership enthusiasm.
• IV behavior on April 24: A gap open followed by rapid IV collapse favors premium-selling structures. Know your exit plan before the open.
• The January lesson: A 275% EPS beat sold off 22%. Intel is not being traded on reported earnings. It is being traded on forward expectations. The guide is the print.

Intel doesn’t need to win the foundry war to have a viable thesis. It needs to show the manufacturing cost curve is improving predictably, that the NVIDIA partnership is producing real products on schedule, and that DCAI is growing faster than CCG is shrinking. Three things. If Thursday delivers on all three — even partially — the stock likely holds. Miss two or more, and the gap between $68 and the $53 analyst consensus closes faster than most are positioned for.

— The Editorial Desk