Tesla bet the Roadster on 6,800 laptop cells — and rode their price curve

When Tesla was founded in 2003, EV orthodoxy held that a car needed large, purpose-built battery cells; the 18650 cylinders from laptops looked like toys. Then AC Propulsion swapped the lead-acid pack in its tzero prototype for laptop-grade lithium-ion 18650s and more than tripled the car's range. Martin Eberhard and Marc Tarpenning — who had already built a product around 18650s at the e-book maker NuvoMedia — licensed AC Propulsion's technology and founded Tesla on the same bet.

Tarpenning described the cells' economics: 'They kept getting better. Every time we looked around, they were another 7% better or 7% cheaper.' A 2006 Tesla paper by Berdichevsky, Kelty, Straubel and Toomre made it explicit: about 6,800 commodity 18650s — 450 kg storing 56 kWh, over 200 miles of range, which no production EV had offered. It credited fifteen years of consumer-electronics market pull.

Safety moved from inside a big custom cell to the system: each of the billions of 18650s made yearly carried its own current limiter, rupture device and steel can, so one failure could not take down the pack. The Roadster shipped in 2008; the Model S pack scaled the bet to 7,104 Panasonic 18650 cells, and the laptop cell became the benchmark every EV battery program was measured against.

Ride a market someone else is funding

The laptop market had already paid the R&D bill: cell density and volume manufacturing improved yearly at a scale no EV program could fund.

Billions of 18650s a year meant mature, high-yield production and predictable quality that any bespoke automotive cell would take a decade to reach.

Small cells localized failures — each of the 6,800 had its own current limiter and rupture device, so one bad cell could not take down the pack.

It decoupled Tesla's fate from its own scale: even building 2,000 cars a year, Tesla bought cells priced for a billion-device market.

If a critical part must keep improving for years, adopt the format another market funds. Inherit its learning curve for free; spend your engineering managing many cheap parts, not perfecting one.

The Roadster's 2008 debut made 'thousands of small cells, managed by software' the working answer to the EV battery question. The Model S repeated the bet at scale with 7,104 cells per pack, and rivals designing large-format cells spent years playing catch-up in cost per kilowatt-hour. Tesla later moved to larger cylindrical formats of its own specifying, but the founding logic — buy the curve the electronics market funds, engineer the system around the commodity part — stayed the company's template, through the Panasonic partnership and into the Gigafactory era.