gōngbǎn 公板 — public board

The system that makes hardware investable.

Most consumer electronics from China are built on a small library of reference designs called gongbans. This briefing explains the system in the language fund managers already understand — and the catalogue shows you the library itself.

Building a product and never made hardware? Start here →

For founders · zero hardware experience is fine

I'm building a product

What a gongban is in plain words, the four steps from idea to production, the mistakes first-timers make, and the words you'll hear. Then the catalogue, 50 product ideas with founder start packs, and the playbook.

Start here →
For investors · fund managers · LPs

I'm an investor

Why hardware has been toxic for fund math, how Shenzhen's gongban system already runs on venture logic, and what a gongban-native portfolio company looks like. Also as a PDF.

Read the briefing ↓

The library, indexed

A static database of Chinese reference designs, built from what chip houses, distributors, module houses and open-hardware communities publish themselves — plus Hans's hand-curated distributor kits with the full design packs. Every entry links to the publisher's own files; nothing is re-hosted except the curated packs.

1,518
Boards
reference designs, dev boards, SBCs, modules
25
Publishers
chip houses, distributors, module houses, maker brands
739
With design files
schematic, PCB, Gerber or BOM published
660
Schematics
entries with a published schematic
66
Solution houses
Shenzhen design houses & ODMs in the directory
3
Full packs hosted
schematic + PCB + Gerber + BOM + spec, downloadable here

50 product ideas built from gongbans

What could a small European team build from this library in a quarter? The GONGBANS inventor combined real catalogue boards into 50 concepts — smart home, energy, installer tools, accessibility, agri, LoRaWAN and more — each with the boards it uses, rough estimates, the EU certification path and the one cheap question that would kill it. Concepts, not products; nothing validated.

Why VCs avoid hardware deals

Hardware has been toxic for fund math. Here's why:

$150K–500K
NRE before revenue
Tooling, molds, certifications — all burned before a single unit ships. Kills fund IRR.
12–18 mo
To first revenue
Dev cycles that eat runway. A SaaS bet gets to market 4x faster with the same capital.
95%
Hardware failure rate
Crowdfunded products. Most die in engineering — a risk VCs can't underwrite or de-risk.

The gongban system eliminates all three risks.

Shenzhen's supply chain already runs on venture logic.

The structure mirrors the fund model investors already know — because both solve the same problem: how to place many bets cheaply and capture outsized returns from the winners.

Once you see this, hardware deals stop looking foreign.

Tier 1 — The LP / PE layer
Your world
Private equity fund
Deploys capital at scale. Backs fund managers. Returns come from portfolio volume across the entire market.
=
Shenzhen equivalent
Chip design house
Rockchip, Allwinner, MediaTek. Makes the silicon. Returns come from chip volume — every unit sold by any brand is revenue.
Tier 2 — The VC layer
Your world
Venture capital fund
Designs the thesis. Invests in 100 startups. Bets that 2–3 will be breakout hits that return the entire fund.
=
Shenzhen equivalent
IDH (design house)
Designs the gongban. Gives it to 100 brands. Bets that 2–3 will be massive sellers that drive huge chip volume back upstream.
Tier 3 — The portfolio companies
Your world
Portfolio startups
100 companies. Each takes the capital, tries to win their market. Most fail. A few become unicorns.
=
Shenzhen equivalent
Hardware brands
100 brands. Each takes the gongban, adds logo + packaging, tries to win their channel. Most stay small. A few become category leaders.

Same portfolio math, different asset class

The IDH's "fund" is a reference design
3–6 months of engineering = one gongban. That's their capital deployment — one bet, spread across 100 brands.
Distribution cost is near zero
Giving the gongban to brand #47 costs nothing — the design is done. Same as writing a check from an already-raised fund.
Returns are power-law
Most brands order 1K–5K units. A few order 500K+. The IDH needs 2–3 hits to return the entire design investment.
The chip company is the LP
Rockchip doesn't care which brand wins. They care about total volume across all 100. They subsidize the IDH to get broad exposure.
"Exits" are volume milestones
When a brand hits 100K+ units, the IDH gets recurring orders. That's the IPO moment of the gongban world.

What changes for your fund

Gongban-native startups flip every metric that made hardware uninvestable:

TraditionalGongban-native
NRE / tooling$150K–500K$0 — already amortized
Time to first unit12–18 months2–4 weeks
Engineering riskHigh — unproven designZero — battle-tested PCB
Minimum viable raise$1M+ seed$50K–100K
Pivot costNew tooling, new cyclePick a different gongban
Where founder winsMust be an engineerBrand, channel, GTM

Why this is your edge

Most Western VCs don't know this system exists. That's the edge.

Hardware deals with SaaS economics
Your portcos launch physical products with near-zero NRE, 3x–5x gross margins, and iterate like software companies. Fund math finally works for atoms.
Information asymmetry
99% of EU/US fund managers have never heard the word gongban. Founders who understand this system have a structural cost advantage that competitors can't see.
Massive TAM, zero crowding
Consumer electronics is a $1T+ market. The gongban layer is how most of it actually gets made — but almost no Western capital is deployed against it.

What a gongban-native portco looks like

The founder you want to back:

1
Picks proven gongban hardware
Curates from thousands of existing, certified, mass-produced designs. No engineering team needed. Start in the catalogue →
2
Adds brand + packaging + channel
Logo, color, retail packaging, EU compliance markings. This is where the margin lives.
3
Owns the customer relationship
D2C, Amazon, retail — they control distribution, pricing, and the post-sale experience.
4
Iterates at software speed
Product not selling? Swap to a different gongban next month. No sunk tooling. No 12-month reset.
Result: a capital-efficient hardware company that a VC can underwrite like a software deal.

Gōngbǎn (公板)

The system that makes hardware investable. Download the full briefing, or go deeper with the book.

↓ Download PDF Get the book