G.fast is an innovative networking technology that promises to deliver very high-speed internet over the vast amount of existing copper infrastructure. By leveraging current copper lines, providers can reach many more customers at much higher speeds with relatively low investment — a prospect that has generated considerable excitement.
Laboratory demonstrations have produced headline-grabbing results. Researchers at Bell Labs achieved record-breaking rates using XG.fast — a related technology — reaching speeds up to 10 Gbps, which is more than 1,000 times faster than typical consumer broadband. While these lab figures are impressive, controlled conditions can differ substantially from real-world deployments, so expectations should be moderated.
To gain a clearer picture of practical performance, BT conducted tests designed to reflect real-world conditions more closely. In trials reported by the company, downstream speeds of around 800 Mbps and upstream speeds exceeding 200 Mbps were achieved over a 19-meter copper run. Extending the copper length to 66 meters — a distance representative of roughly 80% of connections — produced speeds near 700 Mbps downstream and 200 Mbps upstream. These results suggest G.fast can deliver multi-hundred-megabit performance over short copper loops typical of many urban and suburban networks.
Even with G.fast, fiber remains essential. The common deployment model requires fiber to be brought much closer to customers than legacy designs. Traditional approaches like Fibre to the Cabinet (FTTC) and Fibre to the Premises (FTTP) move fiber into cabinets or all the way to homes. G.fast relies on fibre to feed distribution points — a model often called Fibre to the Distribution Point (FTTdp) — while existing copper handles the last few meters into premises. In practice, that means investment in fiber backhaul and new street-side distribution units alongside the reuse of existing copper pairs.
Dr. Tim Whitley, Managing Director of Research and Innovation at BT Group, emphasized the company’s commitment to evaluating the technology: “We see G.fast as a very promising technology with significant potential — that’s why we’re putting our best minds on the case to assess it fully in a purpose-built facility.” He noted BT’s long history of telecommunications innovation and stressed the importance of staying ahead of the curve for the benefit of customers and shareholders.
However, not everyone is convinced G.fast is the best long-term strategy. Dana Tobak, Managing Director of Hyperoptic, cautioned that while G.fast performs well in test environments, rolling it out at scale will demand substantial investment in street cabinets and will likely be most viable in high-density areas. From that perspective, FTTP remains the preferred future-proof solution, and continuing to rely on copper can be seen as delaying the broader adoption of full-fiber networks.
BT has established a purpose-built laboratory at its Adastral Park R&D centre in Ipswich to test G.fast under more realistic conditions and to study its operational characteristics. The company is collaborating with major system vendors, including Adtran, Alcatel-Lucent and Huawei, to explore interoperability, installation practices, and the technology’s limits across a range of real-world scenarios.
G.fast offers a compelling middle ground: a way to achieve very high speeds by reusing extensive copper infrastructure while requiring targeted fiber upgrades. For operators seeking faster services quickly and at lower initial cost than full FTTP rollouts, G.fast could be an attractive option — particularly where short copper runs are common. Nevertheless, its effectiveness depends on local network topology, density, and willingness to invest in distribution infrastructure. As testing continues, operators and regulators will need to weigh those trade-offs alongside long-term plans favoring all-fiber networks.
Do you think G.fast represents the next generation of telecom infrastructure? Share your thoughts in the comments.