Intel Launches 18A-P 'Risk Production': Securing Major Foundry Clients Is the Key
Alina Collins
Intel's 18A-P process has entered risk production, offering 9% more speed at the same power; Apple may test its M7 chip on the node, but whether yields can match TSMC is the real make-or-break.
What does 18A-P actually improve over standard 18A?
18A-P is Intel's enhanced variant of its 18A process. Two headline gains: 9% faster at the same power, or 18% less power at the same speed.
This means → Intel finally has a node whose spec sheet can compete with TSMC's latest on paper.
"Risk production" — trial manufacturing runs before full-scale ramp — is now underway, signaling the process flow is fundamentally working. The next step is pulling outside customers in to validate it.
Apple and Nvidia — who is more likely to place an order?
Nvidia announced a $5 billion investment in Intel, briefly stoking expectations of a major foundry contract — but no clear commitment to use Intel foundry has emerged.
Apple looks more concrete: Bloomberg reported last month that Apple is in early talks with Intel on chip manufacturing. Neither side has commented.
Counterpoint Research analysts believe Apple will likely test its next-generation M7 processor (for Mac and iPad) on 18A-P.
In plain terms = Nvidia wrote a check but hasn't placed an order; Apple hasn't written a check but may send chips to trial first.
Why is yield the real make-or-break?
The analyst quote says it plainly: "Yield matters more than the process node. Matching TSMC's manufacturing consistency is the key."
This means → No matter how good 18A-P looks on paper, the number of working chips each wafer produces is what decides whether customers will move real orders away from TSMC.
This reflects Intel foundry's core dilemma: technology can be caught up on a roadmap, but manufacturing consistency has to be ground out through time and volume.
What does this mean for Intel's finances?
Intel's foundry unit is currently losing billions of dollars per quarter and urgently needs outside customer orders to share fab costs.
Shares rose as much as 3.9% pre-market to $121.60 on the news — but had fallen 8.5% the day before in a broad tech selloff.
In plain terms = the market read it as "good news, but let's fill yesterday's hole first" — far from a trend reversal.
What to watch next?
The first hard checkpoint: Apple's M7 trial-production results on 18A-P. Once yield data is out, Intel foundry's real competitiveness will be impossible to hide.
If yields approach TSMC levels, this becomes a turning point for Intel's foundry ambitions. If the gap is wide, no spec-sheet advantage will be enough to pry away major customers.
This means → Short-term, the trade is headline-driven. Medium-term, what actually prices the stock is yield data. Investors need to wait for hard evidence.
Content is for reference only, not financial advice.