Industry

Industry - Wide-Bandgap Power Semiconductors

Wolfspeed sits inside a narrow, capital-heavy slice of semis called wide-bandgap (WBG) power semiconductors — discrete switches and modules made from silicon carbide (SiC) and gallium nitride (GaN) that handle high voltage and temperature better than ordinary silicon. The WBG device pool is roughly $1.7–2.4B today against an ~$80B silicon power-semi market, with consensus forecasts pointing to $11–14B by 2030 as EV traction inverters, fast chargers, solar/storage and AI-data-center power supplies substitute silicon with SiC. The thing most newcomers miss: this is a materials business wrapped in a device business, the cycle just turned, and a global capacity overhang is now colliding with a softer-than-expected EV ramp.

How This Industry Makes Money

The product is a switch — but one that costs the customer 3–5x what a silicon equivalent costs, justifies the premium by saving 5–10% of system energy losses, and gets designed-in years before it ships. Revenue is overwhelmingly direct B2B with long lead times: design-ins → design-wins → multi-year long-term agreements (LTAs), often with take-or-pay clauses on both sides. Pricing is set per device (dollars per MOSFET) or per wafer (~$1,000–$1,500 for a 200mm SiC bare wafer at recent prices, down sharply from 2022 peaks). Customer concentration is severe: Wolfspeed's top two customers were 37% of FY2025 revenue, and 5–7 year automotive qualification timelines (IATF 16949) entrench the supplier choice once it is made.

Two structural facts dominate the economics. First, fixed costs are enormous. A modern 200mm SiC device fab costs $1.5–3B and runs profitably only above ~60–70% utilization; below that, depreciation and labor turn variable margin negative — exactly what produced Wolfspeed's FY2025 reported gross margin of -16%. Second, the substrate is where the differentiated economics historically sit. SiC crystal growth is slow (centimeters per week, not meters per hour like silicon), defect-prone, and capital-intensive — which is why Wolfspeed, Coherent and ROHM/SiCrystal historically extracted wafer-level margins that device-makers could not. That moat is eroding as Chinese substrate capacity floods in and as IFX/STM/ON in-house their own substrate supply.

Demand, Supply, and the Cycle

The demand engine has three layers stacked on top of each other, and they don't move together. EV traction inverters (the propulsion box that turns battery DC into motor AC) are the largest near-term driver — every percentage point of SiC adoption in a BEV adds ~$300–500 of SiC content per vehicle. Industrial (solar inverters, motor drives, UPS) is steadier and less penetrated. AI data-center power supplies are the newest pull: hyperscalers want SiC and GaN to shrink rack-level losses as cluster power densities climb past 100 kW. The cycle hit in mid-2024 and is still being worked through in 2025–26.

This is a capacity-led trough inside a long-term secular demand story. Capacity was sized in 2021–22 for an EV trajectory that has since moderated, while Chinese substrate entrants compressed the most defensible margin pool. The cycle hits Wolfspeed-style pure-plays hardest because their cost structure is almost entirely fixed; broad-line peers (IFX, STM, ON) absorb the underutilization across larger, more diversified portfolios.

Competitive Structure

The SiC power-device market is a five-firm oligopoly transitioning into something more crowded. Historically Wolfspeed, STM, Infineon, onsemi and Rohm collectively held the overwhelming majority of share — STM as the early traction-inverter leader (Tesla), Infineon as the broad-line automotive king, onsemi as the most aggressive vertical-integrator (acquired GTAT for substrate), and Wolfspeed as the substrate/materials champion. The substrate layer is even more concentrated, but Chinese entrants are reshaping it fast.

Approximate share snapshots, drawn from industry research and inferred from company disclosures rather than a single audited source:

Regulation, Technology, and Rules of the Game

Three rule sets shape returns in this industry: government subsidies that move where capacity gets built, automotive qualifications that lock in who supplies it, and a 200mm wafer transition that resets the cost curve. Each has already determined who survived the 2024–25 trough.

The Metrics Professionals Watch

P/E and revenue growth aren't the binding constraints in SiC. The metrics that move security prices here are operational and cyclical, and they appear earlier in the financials than profit does.

Where Wolfspeed, Inc. Fits

Wolfspeed is the only Western, publicly traded, vertically integrated pure-play in wide-bandgap power semiconductors. That sentence is both the bull case and the bear case. Bull: it owns the entire SiC stack from crystal growth through power module, runs the world's first 200mm SiC device fab (Mohawk Valley, NY), and has been substrate share leader for over a decade. Bear: it is one-fifth the revenue scale of its smallest broad-line competitor, just emerged from Chapter 11 in September 2025, and competes against firms that can subsidize SiC losses with profitable analog and microcontroller franchises.

The simplest mental model for the rest of this report: Wolfspeed is the highest-beta way to express a view on SiC adoption, because it has the most operating leverage (one product family, almost no diversification), the largest 200mm capacity ramp ahead of it, and a clean balance sheet for the first time in five years. That same set of facts makes it the riskiest if the next three quarters of utilization data disappoint.

What to Watch First

The eight industry checkpoints below will tell a reader within one or two quarters whether the backdrop is improving or worsening for Wolfspeed — and where to look for the evidence.