Rewiring AI infrastructure from core to edge

As GenAI shifts from pilots to production, enterprises must rethink infrastructure strategy to meet performance, cost, and compliance demands.

- BY SHUBHENDU PARTH

Nearly 80% of Chief Information Officers (CIOs) in the Asia-Pacific region are expected to adopt edge services from cloud providers by 2027 to meet the performance and compliance needs of generative Al (GenAl) workloads, according to new IDC research commissioned by Akamai Technologies.

The research paper, “The Edge Evolution: Powering Success from Core to Edge,” highlights how the rising adoption of Al is prompting APAC enterprises to reevaluate their digital infrastructure. It finds that centralised cloud models alone are no longer adequate to support the speed, scale, and regulatory requirements of Al inferencing.

According to the IDC Worldwide Edge Spending Guide, public-cloud services at the edge are forecast to grow at a compound annual growth rate of 17% through 2028, with spending expected to reach USD 29 billion.

EDGE SERVICES GAIN MOMENTUM ACROSS APAC

The report reveals that 31% of surveyed enterprises in APAC have already moved GenAl applications into production, while 64% are still in the testing phase. This growing momentum is putting strain on conventional cloud architectures and driving the shift to more distributed models.

Key infrastructure challenges include multi-cloud complexity (49%), changing compliance requirements, which are expected to impact 50% of the A1000 by 2025, unpredictable cloud costs (24%), and latency-related performance issues. These gaps are pushing organisations to integrate edge computing into their infrastructure plans.

India's geographic spread and uneven network quality make edge infrastructure critical for lowering latency and managing costs.