Why Intel Is Heading Into Earnings With Its 18A Story Reversed

Intel reports second-quarter results after the close on Thursday, July 23, 2026, with the stock changing hands around $95.70 and sitting on a gain of more than 160% for the year. Six months ago the setup looked nothing like this. The 18A process node was the single biggest hole in the bull case, and reporting in early July had 18A yields failing to reach commercially attractive levels until late 2026 or even 2027, which combined with AMD's continued share gains in the data center to knock the stock down hard.

Heading into this print, that story has largely inverted. The manufacturing narrative that was the liability is now the reason people own the stock, and Thursday is the first time management has to stand behind it on the record.

- INTC price level: around $95.70

- 2026 performance: up more than 160%

- Earnings date: Thursday, July 23, 2026, after the close

- Peer levels: NVDA around $203.48, AMD around $503.10, MU around $848.95

- Crypto backdrop: BTC around $64,785 and flat

Here is what actually changed in the 18A story, why the ASML milestone is the part worth taking seriously, and what a 160% run means for the bar management has to clear on Thursday.

Where the 18A Story Stood Three Months Ago

18A was supposed to be the node that put Intel back in the same conversation as TSMC. It carried the gate-all-around transistor design and backside power delivery, and the entire foundry pitch depended on it working at volume. When reporting suggested yields would not reach commercially attractive levels until late 2026 or 2027, the pitch fell apart on the spot, because a foundry that cannot promise a customer good die per wafer on a known timeline is not a foundry anyone signs with.

That timing problem arrived alongside AMD taking more data-center share, and the two together explain most of the earlier selloff. Investors were being asked to fund an enormous capital program on the promise that manufacturing would catch up, and the reporting said manufacturing was slipping again. Intel has a long history of exactly that pattern, which is why the market discounted the story so aggressively.

The reason the stock has run this year is that the same story started producing evidence instead of promises.

What Reportedly Changed on Intel 18A Yields

According to industry reporting rather than any figure Intel has confirmed, 18A yields have improved to roughly 85%, up from roughly 65% in the prior quarter. Treat those numbers as secondary-source estimates. Intel has not published them, and yield data is among the most closely guarded information any chipmaker holds.

Taken at face value, that reported level would place Intel second only to TSMC's N2 process, which industry reporting puts near 90%, and ahead of Samsung. The gap between a 65% node and an 85% node is the difference between a process you run for internal parts because you have to and a process you can sell to an outside customer with a straight face. Yield is what determines cost per working chip, and cost per working chip is the only number a foundry customer actually negotiates on.

The honest framing is that the direction of travel looks real and the magnitude is unverified. A qualitative read works fine here. Something clearly improved enough that Intel changed its own behavior around the node, which is usually a better signal than any leaked percentage.

Why Being First on High-NA EUV Is the Strongest Signal

The most solid piece of evidence in the whole reversal did not come from Intel at all. ASML disclosed that Intel was the first chipmaker to reach production qualification on High-NA EUV, the next generation of extreme ultraviolet lithography, and that disclosure came from the equipment supplier rather than from the company being evaluated.

Think of High-NA EUV machines as the printing presses for the most advanced chips of the next decade. Each one runs roughly $350 million, they are the tools that will pattern leading-edge logic for years, and buying one is trivially easy compared with making one work inside a running fab. Production qualification means Intel got the new tool to produce acceptable results in a manufacturing environment, not in a research setting, and did it before anyone else did.

For a company whose entire investment case rests on catching up in manufacturing, that is the most meaningful kind of proof point available, because a supplier verified it. Anyone can put a yield chart on a slide. Very few things in this sector are confirmed by the vendor selling the equipment to all of your competitors at the same time.

Intel also reportedly plans to build Nova Lake in-house rather than outsourcing the work. Companies do not move flagship silicon back onto their own process to be polite about it. That decision is a revealed preference, and it lines up with everything else in the reversal.

The Bear Case Against a Stock Already Up 160%

A 160% year is not a neutral starting point, and this is where the article stops being a bull note.

Start with the obvious. A move of that size means a great deal of good news is already reflected in the price, so Thursday is being graded against expectations that have climbed all year rather than against the depressed expectations of six months ago. Beating the old bar does nothing. Management has to beat the bar the rally itself created.

Then the specifics. The yield figures are not company-confirmed, which means a management commentary that comes in vaguer than the reported numbers would read as a soft denial. Foundry customers still have to actually sign, and a production-qualified tool plus an improved node is a precondition for winning external business rather than the win itself. Announced intent and signed volume are different things in this industry, and Intel has a documented history of announcing progress that later slips.

Here is the part most retail traders underweight. When a stock has already run, the asymmetry inverts. Good news gets you a modest move, and anything ambiguous gets you a sharp one in the other direction, because the marginal holder bought recently and has no cushion.

What the Semiconductor Selloff Adds to the Setup

This print also lands days after Moonshot AI's Kimi K3 release pushed the PHLX Semiconductor Index into a bear market on Friday, July 17, 2026, as the market repriced how much AI capex the efficiency gains actually justify. That repricing hit the whole complex, from NVIDIA at the top of the accelerator stack down through the memory names like Micron, and it is covered in depth separately alongside the wider AI semiconductor stock rotation.

The wrinkle for Intel is that a capex-doubt selloff and a manufacturing-turnaround story pull in opposite directions. If the sector is repricing on the assumption that less compute gets bought, that is a demand problem, and no amount of yield improvement fixes a demand problem. Intel's server business is exposed to exactly that question, which is why demand durability sits next to yields on Thursday's watch list.

Crypto, for its part, has stayed out of the way. BTC is holding near $64,785 and flat, so the Bitcoin tape is not adding or subtracting risk appetite from the semiconductor trade this week.

What to Watch on Thursday July 23

Three things decide how the print lands, and they are not equally weighted.

Yield commentary is the one that matters most, because it either validates the reported numbers or leaves the entire reversal resting on secondary sourcing. Note that management is not obligated to give a figure, and a refusal to quantify is not automatically bad news. It is, however, exactly the ambiguity a stock up 160% handles poorly. The Intel investor relations page carries the release and the call, and ASML's EUV lithography documentation is the reference for what production qualification on High-NA actually involves.

Frequently Asked Questions

When does Intel report Q2 2026 earnings?

Intel reports second-quarter 2026 results after the market close on Thursday, July 23, 2026, with the conference call following the release. The call matters more than the numbers this quarter because management commentary on 18A is the swing factor.

Are Intel's 18A yields confirmed at 85%?

Intel itself has not confirmed that number anywhere in its public disclosures. The roughly 85% figure, up from roughly 65% in the prior quarter, comes from industry reporting rather than from the company, and Intel has never published yield data of that kind. Treat it as a directional estimate until management addresses it on the call.

What does High-NA EUV production qualification mean?

It means Intel has demonstrated that the next-generation lithography tools produce acceptable results inside a real manufacturing flow rather than a lab. ASML disclosed that Intel reached this first, which makes it one of the few claims about Intel's manufacturing progress that a third party verified.

Is INTC stock a buy before earnings?

That depends entirely on how much of the turnaround you think a 160% year already reflects. The manufacturing evidence has genuinely improved, but buying into a print after a move that size means accepting an asymmetric setup where ambiguity costs more than confirmation pays.

The Bottom Line

The 18A story flipped from liability to thesis in roughly six months, and the ASML High-NA disclosure is the anchor because a supplier verified it rather than Intel claiming it. That is a real change in the quality of evidence, and it explains why the stock trades near $95.70 instead of where it sat in the spring.

What it does not do is lower the bar. Confirm the reported yield trajectory on the call and name a foundry customer, and the reversal gets a second leg. Deflect on yields, or guide server demand softer into the AI-capex repricing that just put the semiconductor index into a bear market, and a stock up 160% has a long way to fall before it finds buyers who were not already positioned.

The setup is straightforward. Intel has spent this year being rewarded for evidence it did not have to provide itself. Thursday is the first time it has to.

This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency trading involves substantial risk. Always conduct your own research before making trading decisions.