In 1990, three years before the efficient blue LED existed: LED technology already measured 32.5%/yr against incandescent lighting's 8.6%. Seventeen years into that signal, the industry's biggest name bet on building a better filament.
Every rate reading below is what the platform computes from patents public in that year.
White LED lighting does not exist yet: LEDs are indicators, not illumination. The improvement rate already reads four times the incandescent bulb.
At Nichia, Shuji Nakamura demonstrates the efficient blue LED, the breakthrough that makes white LED light physically possible.
LED filings pass incandescent, a category the filament had led for a century. The consensus arrives twelve years after the rate.
GE announces its High-Efficiency Incandescent programme, targeting twice the efficiency by 2010. That year LED reads 47%/yr and files 7,899 families against incandescent’s 2,625.
The U.S. Energy Independence and Security Act sets efficiency standards traditional incandescent bulbs cannot meet; the EU follows in 2009.
Twenty months after announcing HEI, GE suspends it, in its own words, to place greater focus on LED and OLED. The rate had said so for eighteen years.
Akasaki, Amano and Nakamura win the Physics Nobel for the blue LED that, in the committee’s words, lit the world.
The business Thomas Edison founded is sold to smart-home company Savant, the end of the incumbent that once defined the industry.
Thirty-six years into the series, LED still reads 37.7%/yr. The filament reads 11.
Defending the filament was not stupidity, it was reasoning from the catalogue instead of the curve.
In 1990 this was simply true: no white light, wrong price, wrong brightness. The rate measured the trajectory of a semiconductor, not the catalogue of the day.
Early LED lamps cost dozens of times more than a bulb. A technology improving at 35%/yr closes a 50× cost gap in roughly a decade, and did, on schedule.
Warmth, dimming, familiarity, real preferences that made bulb bans politically contested. Preference delayed nothing except the incumbents’ own transitions.
GE’s HEI programme was a serious engineering bet on doubling incandescent efficiency. Against a competitor compounding at 4× the rate, doubling once was never going to be enough.
We forecast the improvement rate from two citation metrics, cycle time and knowledge flow. Both were visible the whole time.
LED’s cycle time ran five to six years against the incandescent bulb’s twelve to fifteen, more than twice the iteration speed, sustained for three decades. Its knowledge flow tripled the filament’s: later work kept building on LED inventions and stopped building on filament ones.
Provenance. Every patent carries a filing date, so the record can be replayed: every improvement rate on this page uses only the patents that were public at the time. No hindsight enters the series. Chart data: GetFocus platform exports, August 2026. The historical facts below come from the public record.
Bring us one live technology decision. Within 48 hours you get a free preliminary Decision Brief: a first-pass radar, improvement rates and suggested triggers.