# Cambridge GaN Devices Financial Model

Fabless semiconductor startup commercialising self-protected GaN power transistors (HiGaN) that replace silicon and standard GaN with higher efficiency and no external protection circuitry.

- Canonical: https://finamodel.com/startups/cambridge-gan-devices
- Excel download: https://finamodel.com/startup-models/cambridge-gan-devices.xlsx
- Category: Hardware/Deep-tech
- Model type: Marketplace / GMV
- Funding round: Series A
- Funding: $95M
- Founded: 2021
- Geography: UK (Cambridge HQ); EU partnerships (NL, DE); targeting global OEM markets.
- Customer: B2C

## About the company

Cambridge GaN Devices commercialises self-protected gallium-nitride power transistors for more efficient power electronics. Its fabless HiGaN products are designed to replace silicon and standard GaN components without requiring external protection circuitry, potentially simplifying customer designs and improving efficiency.

The important commercial milestone is a design win: customers must qualify the component and then incorporate it into products that ship at scale. Revenue therefore arrives after a potentially long engineering and validation cycle, while manufacturing depends on wafer supply, packaging, testing, and concentrated customer or channel relationships.

Model qualified design wins, time to production, unit shipments, ASP, wafer and packaging cost, test yield, and customer concentration. Include applications engineering, R&D, foundry commitments, sales, and inventory costs. Win rate, qualification timing, shipment ramp, pricing, gross margin, supply yield, and repeat designs should determine scenarios.

## What's included

- 5-year monthly revenue build with stage-appropriate growth assumptions
- Full P&L, headcount plan, and operating-expense schedule
- Cash-flow statement, runway, and burn-rate tracking
- Valuation via exit multiple with a DCF cross-check
- Returns analysis with MOIC and IRR
- Unit economics including CAC, LTV, payback, and cohort retention

## Product & value proposition

- Product: CGD NOR-02 GaN transistor (visible on cover slide) - a chip with integrated on-chip IC ("HiGaN intelligence") that removes external protection/gate-driver components needed by standard GaN devices.
- Key claims: up to 99% efficiency; switching frequency up to 100× faster than silicon; no additional gate driver, safety, or reliability components required.
- Differentiator vs. standard GaN: standard GaN requires extra external components that degrade efficiency and increase BOM cost; CGD integrates all protection on-chip.
- Markets served: consumer power supplies/chargers, industrial LED drivers, data centre power supplies (OPEX reduction ≥10%, CAPEX reduction up to 10%), automotive on-board chargers and DC-DC converters (50% smaller OBCs).

## Market

- TAM: Power semiconductor market = $45 billion.
- SAM (GaN devices sub-market): Predicted to reach over $1 billion by 2026.
- Chart read (slide 7 image): GaN power device market ~$0.05B in 2017, growing to ~$0.2B by 2020, ~$0.6B by 2022, ~$1.1B by 2025, ~$1.75B by 2028. Segments: PSU Computing/Datacenters (largest, dark blue), PSU Consumer/Chargers (mid blue), PSU Other/Lighting (grey), EV/HEV (orange), Other (light blue).
- Implied CAGR (2017–2026, GaN sub-market): ~$0.05B → >$1B over ~9 years ≈ ~40% CAGR.

## Revenue model

- Fabless model: CGD designs chips; outsourced fab/packaging supply chain "in place" as of 2019.
- Revenue stream: Component sales (transistors) to OEM/ODM customers - price per unit × volume shipped. No SaaS, licensing, or subscription layer described in deck.
- Channels: Direct B2B engagement ("Customers' engagement" noted 2020, "1st Design-In" noted 2021).

## Traction & metrics

- 2016: Founded; 20+ years team GaN experience; 1st prize Cambridge Enterprise competition.
- 2017: First Innovate UK (iUK) project awarded (lead-partner).
- 2018: Second iUK project awarded.
- 2019: BEIS project awarded (green technologies); $2.5M seed fund raised; +4 patent applications; full supply chain in place.
- 2020: $12M European (GaNext/Penta EUREKA) project awarded (lead-partner, 13 partners, 3 countries); customers' engagement commenced; +7 patent applications; engineering samples delivered.
- 2021 (current at time of deck): 1st Design-In achieved; raising $9.5M Series A; IP portfolio = 26 patent applications.

## Competition / moat

- Competitors (implied): Standard GaN players (Infineon, NXP, Vishay, On Semi - all cited as team's prior employers implying familiarity with competitive landscape); silicon incumbents (low frequency, lower efficiency).
- Moat: 26 patent applications covering HiGaN integrated protection technology; Cambridge University affiliation and brand; team with direct insider knowledge of competitor products; CTO (Prof. Udrea) with 150+ patent applications and two prior fabless exits (CamSemi → Power Integration; Cambridge CCS → AMS).
- Competitive positioning chart (slide 5): CGD claims highest efficiency (~99%) at highest frequency (1 MHz+), versus standard GaN (high efficiency, hard to use) and silicon (easy to use, lower efficiency/frequency).

## Team & funding ask / use of funds

- Dr. Giorgia Longobardi - Founder & CEO; 11 years GaN power device experience; NXP, Infineon, Vishay; led GaN team at Cambridge Engineering Dept.; 12 patent applications.
- Prof. Florin Udrea - Founder & CTO; Cambridge Professor; 150+ patents; prior exits: CamSemi (to Power Integration), Cambridge CCS (to AMS), Camutronix.
- Zahid Ansari - VP Operations; ex-CamSemi VP Operations; scaled to 500,000 ICs/day, 1B+ ICs total.
- Andrea Bricconi - VP Business Development; 20+ years power semiconductors; ex-International Rectifier, Vishay; co-started GaN program at Infineon.
- Funding ask: $9.5M Series A.
- Prior funding: $2.5M seed (2019) + substantial non-dilutive public grants (iUK ×2, BEIS, $12M EU GaNext lead).
- Use of funds: Not explicitly stated. Implied: commercialisation phase - market entry, scaling from engineering samples to volume production.

---

## Recommended financial model

- **Archetype + why:** Fabless semiconductor revenue P&L + cash burn model. CGD is pre-revenue, selling discrete components (transistors) to OEMs at per-unit ASPs with volume ramp. This is not SaaS, not marketplace, not DTC. The right structure is: unit volume ramp × ASP → revenue; COGS via outsourced fab (wafer cost + packaging + test); gross margin expansion as volumes scale; R&D + SG&A opex; burn rate and runway vs. Series A proceed. No balance sheet complexity needed yet; a 3-statement model is premature pre-revenue - a P&L + cash model suffices.

- **Forecast horizon & granularity:** 5 years (2021–2025), quarterly for years 1–2, annual for years 3–5. Rationale: first commercial revenues likely 2022 (design-in → production ramp takes 12–18 months); Series A runway and next funding gate are the near-term critical outputs.

- **Key drivers & assumptions:**

| Driver | Value / Source |
| -- | -- |
| First revenue year | 2022 |
| Year 1 unit shipments | ~50k–200k units; early ramp with 1–2 anchor customers |
| ASP per unit | $3–$8; GaN transistors in consumer/industrial class; no deck data |
| ASP erosion | ~5–10% per year as competition intensifies |
| Unit volume CAGR (yrs 1–4) | ~100–150% (fabless semi ramp profile); design-wins cascade |
| Gross margin (at scale) | 45–55%; fabless semi benchmark (no fab capex; outsourced wafer/pkg/test costs) |
| Gross margin year 1 | 25–35%; low volumes, high per-unit fab cost |
| R&D opex | $3–4M/year; team salaries + IP maintenance + tape-out costs; partially offset by grant income |
| SG&A opex | $1.5–2.5M/year; small commercial team (CEO/VP BizDev + hires) |
| Grant income (non-dilutive) | ~$1–2M/year ongoing; CGD has strong track record but amounts unconfirmed post-Series A |
| Series A proceeds | $9.5M |
| Seed remaining | partially spent by 2021; treat as near-zero by close of Series A |
| Headcount | 15–25 FTEs by 2022, scaling to 40–60 by 2025 |
| GaN market growth rate | ~40% CAGR through 2026 |
| CGD market share (2025) | 0.5–2% of GaN sub-market; conservative given early entrant status |
| Target segments priority | Data centres → consumer chargers → automotive OBC |

- **Scenarios (Base / Bull / Bear):**
  - **Base:** First volume revenues in Q2 2022; 2 design-wins by end 2022; gross margin reaches 45% by 2024; Series A funds 24–30 months of runway; Series B needed ~2023.
  - **Bull:** Design-win acceleration (Infineon/Signify GaNext partners convert to customers); automotive OBC design-in by 2023; gross margin 50%+ earlier due to competitive ASP holding.
  - **Bear:** Design-in → production ramp delayed 6–12 months (customer qualification delays are common in semi); single design-win stalls; Series A runway shrinks to 18 months; bridge round required.
  - Flex variables: time-to-revenue, design-win count, ASP, gross margin ramp pace, grant income continuity.

- **Required sheets / outputs:**
  1. Assumptions dashboard (all drivers in one place, tagged /)
  2. Revenue build (units × ASP by segment: consumer, industrial/data centre, automotive)
  3. COGS build (wafer cost + packaging + test per unit; scale curve)
  4. P&L (revenue → gross profit → EBITDA → net loss)
  5. Cash flow / burn (operating cash + grant receipts + Series A proceeds; monthly/quarterly runway)
  6. Headcount & opex schedule
  7. Scenario toggle (Base / Bull / Bear on key assumptions)
  8. KPI summary (units shipped, ASP, gross margin %, cash runway in months)

## Frequently asked questions

### Is the Cambridge GaN Devices financial model free?

Yes. The Cambridge GaN Devices model is a free Excel download with live formulas.

### Can I change the assumptions?

Yes. The workbook is editable and its live formulas recalculate when assumptions change.