By Light Reading
Publication Date: 2026-04-29 08:00:00
The telecom industry shines at raising and dashing hopes. The 5G standard launched in 2019 promised remote surgery and billions of connected objects, and for some unfathomable reason was supposedly needed in self-driving cars. Seven years later, its main use case is speeding up TikTok downloads. Undaunted, 6G’s salespeople are already hitching the embryonic new standard to AI. Thinking humanoids, not bus-riding teens, could be its main customers. But for Laurent Leboucher, the no-nonsense group CTO of Orange, AI’s real attraction today is in potentially making his radio access network (RAN) more efficient.
That explains Orange’s recent new tie-up with Nokia and Nvidia. Orange already uses Nokia’s purpose-built 5G products at mobile sites in France and other countries. It has previously had no obvious need for Nvidia, the giant US chipmaker whose graphics processing units (GPUs) train AI’s large language models. But Nokia and Nvidia became inextricably joined last October when the chipmaker bought a 3% stake in the RAN vendor with an investment of $1 billion. Nokia is now designing RAN software to work on GPUs.
Leboucher’s interest in all this stems partly from his apparent concern about the cost of producing custom silicon – the application-specific integrated circuits (ASICs) of a purpose-built 5G network. “It creates an opportunity to bring a general-purpose chipset instead of an ASIC implementation,” he told Light Reading at last week’s FutureNet World event in London….
