Energy efficiency, not raw compute, will define the next phase of edge AI: IISc’s Viveka Konandur Rajanna

BENGALURU: The biggest challenge in taking artificial intelligence deeper into edge devices is no longer simply increasing compute performance, but delivering that intelligence within extremely constrained energy budgets, according to Dr Viveka Konandur Rajanna, Assistant Professor, Indian Institute of Science (IISc).

Speaking during a session on “Intelligence at the Edge: Designing AI Within the Energy Budget” at the ETElectronicsWorld Design & Verification Summit 2026 in Bengaluru, Rajanna said AI workloads are advancing considerably faster than the battery, memory and computing capabilities available to many edge systems.