Uploaded March 2026 | Updated September 2026, 2 weeks ago
At MWC 2026, Chris Bergey, EVP, Edge AI Business Unit, Arm, joins Jessica Payne from the Raine Group, Ishwar Parulkar, CTO of Telecommunications Vertical at AWS, and Quang Hieu Le, CIO for Viettel Networks, on a GSMA panel to explore one of the most critical challenges facing the industry today - scaling AI across data, compute, and energy.
As AI adoption accelerates, the pressure on infrastructure is intensifying. This panel examines where the real bottlenecks are emerging, from power constraints in hyperscale data centers to the growing need for efficient compute and distributed AI at the edge.
Chris Bergey also shares Arm’s perspective on enabling the next phase of AI through power-efficient architectures, custom silicon, and scalable compute across cloud and edge environments. The discussion highlights the evolving role of telecom networks, the shift from training to inference, and the scale of investment required to support global AI demand.
00:00 Introduction
03:26 The biggest bottlenecks in AI today
07:57 How to alter energy footprint for AI
12:48 How are telcos uniquely positioned in the AI era
19:15 How does AI impact the energy equation as it moves to the edge
21:30 What AI innovations are tackling the data, energy, and compute question
23:11 AI Insights and opportunities that are unique to Vietnam
28:08 What factors to look at when investing in innovation
Copyright/courtesy to GSMA/MWC.
Stay connected with Arm: arm.com
At MWC 2026, Chris Bergey, EVP, Edge AI Business Unit, Arm, joins Jessica Payne from the Raine Group, Ishwar Parulkar, CTO of Telecommunications Vertical at AWS, and Quang Hieu Le, CIO for Viettel Networks, on a GSMA panel to explore one of the most critical challenges facing the industry today - scaling AI across data, compute, and energy.
As AI adoption accelerates, the pressure on infrastructure is intensifying. This panel examines where the real bottlenecks are emerging, from power constraints in hyperscale data centers to the growing need for efficient compute and distributed AI at the edge.
Chris Bergey also shares Arm’s perspective on enabling the next phase of AI through power-efficient architectures, custom silicon, and scalable compute across cloud and edge environments. The discussion highlights the evolving role of telecom networks, the shift from training to inference, and the scale of investment required to support global AI demand.
00:00 Introduction
03:26 The biggest bottlenecks in AI today
07:57 How to alter energy footprint for AI
12:48 How are telcos uniquely positioned in the AI era
19:15 How does AI impact the energy equation as it moves to the edge
21:30 What AI innovations are tackling the data, energy, and compute question
23:11 AI Insights and opportunities that are unique to Vietnam
28:08 What factors to look at when investing in innovation
Copyright/courtesy to GSMA/MWC.
Stay connected with Arm: arm.com
