Not like conventional public clouds, these approaches are sometimes constructed from the bottom as much as deal with the distinctive calls for of contemporary AI infrastructure. This implies high-density GPU configurations, liquid cooling methods, and energy-efficient designs. Extra importantly, they permit enterprises to shift to possession fashions or shared sources that reduce prices over the long run.
Betting on the fallacious enterprise mannequin
Public cloud suppliers are positioning themselves because the pure dwelling for constructing and deploying AI workloads. Naturally, the main focus at AWS re:Invent 2024 was once more on generative AI and the way the AWS cloud helps generative AI options. Early-stage AI experimentation and pilots have pushed a short-term spike in cloud income as organizations flock to hyperscalers to coach advanced fashions and quickly take a look at new use instances.
Coaching AI fashions on public cloud infrastructure is one factor; deploying these methods at scale is one other. By betting on AI, public cloud distributors are relying closely on consumption-based pricing fashions. Sure, it’s simple to spin up sources within the cloud, however the cracks on this mannequin have gotten more durable to disregard. As firms shift from experimentation to manufacturing, long-term, GPU-heavy AI workloads don’t translate into price efficiencies.
