The Foundry / Customer Partnership Pitch: Selling Capacity When the Wafer Is the Product

The Foundry Deck's Real Audience

The fabless chip company sits across the table with a taped-out design, a wafer start target, and a quiet concern no one will state aloud: is this foundry going to treat us like a volume customer or a development partner? This is not a commodity procurement. The fabless partner is committing its next-generation product — potentially its entire revenue horizon — to a process node that may not hit target yield for eighteen months. The stakes are asymmetrical. The foundry loses a line item. The fabless company loses a market window, an entire SoC generation, and the trust of its own customers. A deck that opens with capacity tables and price-per-wafer figures misses the actual decision: can this foundry’s process team learn faster alongside our design team than the competition’s can? That question is operational, not transactional. The deck must prove the foundry understands the fabless partner’s product timeline better than the fabless partner’s own operations team does — because if that trust doesn’t land in the first three slides, the rest of the presentation becomes a negotiation over details that will never matter.

Why This Deck Is a Different Animal from Other Hardware Pitches

A semiconductor foundry pitch does not resemble a component supplier deck or a contract manufacturer proposal. Those relationships involve spec sheets and purchase orders. This one involves process development kits, mask sets, and a shared dependency on defect density curves that neither party fully controls. The fabless partner’s engineering team is evaluating a technology partnership that will consume millions in non-recurring engineering costs before a single functional die ships. The foundry’s manufacturing team is evaluating whether to dedicate clean-room capacity that could otherwise serve higher-margin customers. Neither party sees the full picture alone. The real external force here is the capital intensity of advanced-node fabrication. A single 300mm wafer at a leading-edge node like N3 or N2 carries a cost that makes automotive supply chains look low-stakes. TSMC’s 2023 capital expenditure of $30 billion is not an abstract data point — it is the competitive backdrop. The deck must show the fabless partner that the foundry can make that kind of investment stick to a specific process roadmap without passing every risk downstream. That means talking about yield learning curves, defect density targets, and bin-sort distributions in the same room where someone is also talking about payment terms.

Building the Deck: Sequence, Substance, and the Capital Project Arc

This deck follows a Capital Project Approval Arc because that is what the fabless partner’s decision committee is actually running. The fabless company’s chief technology officer and chief financial officer will evaluate this proposal the same way they evaluate a new tape-out flow or a new EDA tool investment — through a lens of risk-adjusted return on engineering spend, not marketing enthusiasm. The sequence must respect that frame. Start with a Process-Competence Slide that names the specific node, the target customer application, and the reference process data from a known competitive baseline. Do not lead with capacity numbers. Lead with the engineering validation that makes the capacity relevant. Second slide: the Capacity & Allocation Model. The fabless partner needs to see committed wafer starts per quarter, ramp-up timelines, and bin-sort yield targets stated as ranges, not promises. Third slide: the Financial Partnership Structure. This is where the deck breaks from a standard sales proposal. Show the NRE amortization curve, the wafer price trajectory as yield improves, and — critically — how risk is shared if the process falls behind the yield learning curve. Fourth slide: IP Protection Framework. This is the trust slide. Detail the firewalls, the audit rights, the restrictions on process reuse. The fabless partner’s quiet fear is that the foundry will take the design innovations from this partnership and bake them into the standard cell library offered to the next customer. Address that openly. Fifth slide: the Joint Development Roadmap. Lay out the engineering milestones — PDK releases, tape-out support milestones, characterization targets — as a shared timeline. End with a Commitment Slide that asks for a specific next step: a process qualification run, a joint review of the PDK, or a definitive agreement with a launch date. Every slide answers one question the Capital Project committee is already asking.

When the Engineering Detail Exceeds Internal Capability

The gap between a generic foundry pitch and a winning one is almost always a matter of technical depth. Most foundry marketing decks use the same semiconductor imagery and the same ‘advanced node capabilities’ boilerplate. The fabless partner’s technical staff sees through that in seconds. What they need is a deck that reads like an engineering brief: process cross-sections drawn to scale, electromigration data presented as actual measurements, and a yield ramp projection that accounts for defect Pareto charts from the foundry’s existing runs on similar nodes. Very few internal teams have both the semiconductor domain fluency and the presentation design discipline to produce that. The presentation design shop that built the standard corporate overview deck is the wrong resource for this. A foundry partnership deck requires someone who can read a transistor cross-section, understand why a front-end-of-line defect matters more than a back-end-of-line defect for this specific node, and then translate that technical hierarchy into a visual flow an executive committee can follow. Presentation Gurus works with semiconductor teams precisely at this intersection of process engineering depth and decision-making clarity. The deliverable is a work order for a deck that makes the fabless partner’s CTO think ‘this foundry understands my product timeline better than I do’ — not a sentence about being ‘a trusted partner.’

The SCQA Engine That Makes the Partnership Stick

This deck type’s storytelling structure is SCQA — Situation, Complication, Question, Answer — and the reason matters for how the fabless partner engages with the material. A fabless chip company does not arrive at the meeting unsure of its own product. The situation is already settled: the design exists, the market window is known, the competitive landscape is mapped. The foundry pitch cannot pretend to be a blank-sheet discovery. The complication is that the foundry’s process capability, capacity commitments, and IP protections are unproven from this partner’s perspective. The deck’s job is to surface that complication directly — not to smooth it over — and then answer it with specific engineering, commercial, and legal responses. The fabless partner’s attention pattern is skeptical and detail-forward. They skip narrative flourishes. They double back on yield assumptions. They tune out during standard corporate ‘mission and vision’ slides. SCQA works because it mirrors their own engineering problem-solving discipline: here is the known situation, here is the specific risk we both need to solve, here is the question that risk creates, and here is our answer. The structure demonstrates partnership by walking through a shared problem to a shared solution.