Tutorials Streaming

Charles Kao: The Loss Was Impurities, Not Physics

By Infinite Creation · Profile

Historical photograph of Charles K. Kao.

At a Glance

Charles Kuen Kao (1933-2018) demonstrated in 1966 that the attenuation of light in glass fibre was caused by impurities rather than by any intrinsic property of glass, and specified the purity required for practical long-distance optical communication. He shared the 2009 Nobel Prize in Physics. Every submarine cable and fibre network in the world rests on that finding.

Lived: 1933-2018, China, worked in Britain, the US and Hong Kong
Anchor year: 1966, the fibre attenuation paper
Strand: canon

Career

The problem as it stood

Optical fibres of the early 1960s lost about 1,000 decibels per kilometre: a signal was gone within metres. The consensus was that this was intrinsic to glass, making optical communication a non-starter over any useful distance.

What he actually did

Kao, at Standard Telecommunication Laboratories in Harlow, analysed the sources of loss and showed the dominant term was absorption by impurities, particularly iron and other transition metals, not scattering intrinsic to the material. He specified the target: below 20 dB per kilometre, which he argued was achievable with sufficiently pure silica.

How it worked

Separating intrinsic loss from extrinsic loss changes the problem from a physical limit into a purification target. Corning reached 17 dB/km in 1970. Modern fibre achieves around 0.2 dB/km, meaning light travels tens of kilometres before needing amplification.

What it made possible

The physical layer of the modern internet. Intercontinental submarine cables, backbone networks and fibre to the home all depend on it, and streaming video, which moves far more data than broadcasting ever did, is only economic because fibre capacity is effectively enormous.

What happened to him

He became vice-chancellor of the Chinese University of Hong Kong, shared the 2009 Nobel Prize, and was diagnosed with Alzheimer's disease around the time of the award. He died in 2018.

Where the credit landed

Kao is properly credited and Nobel-recognised, and his contribution is frequently described as inventing fibre optics, which it was not: fibres existed and Narinder Singh Kapany had demonstrated image transmission through bundled fibres in the 1950s. Kao's contribution was the diagnosis that made low-loss fibre a target worth pursuing, which is a different and arguably harder kind of insight.

Known For

He reframed a supposed physical limit as a purification problem.
Modern fibre reaches ~0.2 dB/km, against ~1,000 dB/km in the early 1960s.
He did not invent optical fibre. He showed low-loss fibre was possible.

Knowledge Base

What You'll Learn

The most valuable result here was a diagnosis, not a device.

Why streaming needed fibre

Broadcasting sends one signal to everyone, so its cost is independent of audience size. Streaming sends a separate stream to every viewer, so cost scales with the audience: an approach that is absurd unless bandwidth is extraordinarily cheap. Fibre is what made it cheap enough. The economic inversion at the end of this corpus, from one-to-many to one-to-one delivery, rests directly on Kao's finding.

Where This Fits

Low-loss fibre carries signals between the ends of the chain traced in A History of Broadcast Technology.

FAQ

Q: Did Charles Kao invent fibre optics?
A: No. Optical fibres already existed and had been used for image transmission. Kao showed in 1966 that their high loss came from impurities rather than from glass itself, and specified the purity needed for practical long-distance communication.

Q: How much better is modern fibre?
A: Early 1960s fibre lost roughly 1,000 decibels per kilometre. Corning reached 17 dB/km in 1970 and modern fibre achieves about 0.2 dB/km.

Vocabulary

Optical fibre. A thin glass strand carrying light along its length by reflecting it back inside. Carries far more information than a copper wire of the same size, which is why long-distance networks use it.

Attenuation. The loss of signal strength over distance, measured in decibels per kilometre. Determines how far a transmission travels before it needs regenerating, and it was the barrier that made glass seem unusable.

Total internal reflection. The effect where light striking a boundary at a shallow enough angle reflects entirely rather than passing through. What keeps light inside a fibre instead of leaking out of its sides.

Bandwidth. How much information a channel can carry per second. Optical fibre offers vastly more than electrical cable, which is what made streaming video to millions of homes practical rather than theoretical.

Sources

The paper was checked against its publication record. Checked 20 August 2026.

K. C. Kao and G. A. Hockham, Dielectric-Fibre Surface Waveguides for Optical Frequencies. Proceedings of the Institution of Electrical Engineers, volume 113, issue 7, pages 1151 to 1158, July 1966. This is the paper the 1966 anchor year above refers to. The record names George A. Hockham as co-author alongside Kao.