What Nvidia's survey says about AI in telecom

AI giant Nvidia released its second annual survey of AI usage in the telecom industry. The survey's results highlight both the growing interest in AI technology among executives in the telecom industry as well as Nvidia's high-profile role as perhaps the most important company in the AI sector in general.

Among the survey's findings: Telecom companies are increasingly turning to a hybrid cloud approach – one that uses a mixture of cloud and on-premises infrastructure – in order to run their AI projects. A large number are using AI for customer care services, but a significant number are also looking to apply the technology to their network operations, including the radio access network (RAN). And telecom companies are increasingly struggling to find enough data scientists to head up their AI efforts.

Nvidia said it derived its findings from a survey of 400 "telecom industry professionals" that was conducted between October and December 2023.

Nvidia's outsized role

Of course, AI has arisen as a major topic in the telecom industry and the wider tech industry. Services like ChatGPT and Microsoft's Copilot have galvanized interest in the ways AI can improve existing systems or create new ones. As a result, companies like Google, Amazon and Microsoft are investing billions of dollars into the core data center architecture necessary to run high-performance AI services.

And Nvidia is central to that investment, considering the company's graphical processing units (GPUs) are viewed as essential to running most advanced AI services. As a result, investors have been flocking to Nvidia, sending the company's stock to sky-high levels. Indeed, Nvidia this week passed both Amazon and Google's Alphabet in total market capitalization, making Nvidia the third most valuable company in the US. Nvidia is scheduled to release its quarterly results early next week, and it's possible the company's shares could climb even higher.

That's why Nvidia's view into the state of AI in telecom is important. The company first conducted a survey of AI in telecom last year, and this year it released its second "State of AI in Telecommunications" edition.

"We're starting to get some actual, real numbers, meaningful numbers around the impacts (of AI)," Nvidia's global head of business development for telco, Chris Penrose, told Reuters. Penrose is widely known in the US telecom industry as the executive who represented AT&T's Internet of Things (IoT) business for years, before moving to IoT player FogHorn in 2020. According to his LinkedIn profile, Penrose joined Nvidia in 2022.

The findings

According to Nvidia's new survey, "AI adoption, especially generative AI, is growing, with a notable increase in AI enthusiasm at the executive level [in telecom]. In addition, compared to last year's report, there's better clarity on the role of AI in the industry, the number of AI use cases is increasing, and the return on investment (ROI) from AI is better understood."

The survey found that 41% of respondents are implementing or using AI, while only 10% are not using the technology. Respondents are also primarily interested in using the technology to improve customer experiences, but a significant number – more than 30% – also want to use it to reduce costs.

In Nvidia's 2022 survey, respondents named their inability to find a suitable return on their AI investment as their biggest challenge. But this year respondents said their biggest challenge was finding suitable data scientists to implement their AI strategy.

In terms of AI networking and operations, 34% said they hoped to apply AI to their network systems, including the RAN. But that's down from 38% in 2022.

Finally, 49% of respondents said they hope to use a hybrid cloud approach to AI next year. 33% said they would take a cloud approach, and just 18% said they would take an on-premises approach.

"The role of AI in the telecom industry is deepening," the Nvidia report summarized. "Insights from industry respondents make it clear that the status and impact of AI are accelerating."