AWS has long been the top cloud provider worldwide, but in recent years, it has lagged behind its closest rivals in AI.
That’s changing, as the vendor steps up its game in generative and agentic AI.
The tech giant marked its emergence as a bona fide player in the burgeoning global AI market at its annual AWS re:Invent conference this week, releasing a barrage of new AI systems and tools, including frontier agents, the Nova Forge agent-building service, Nova foundation models, AI factories and more advanced and faster versions of its Trainium AI chips.
In a special breaking news edition of the Targeting AI podcast from AI Business, R “Ray” Wang, founder and analyst at Constellation Research, broke down AWS’s AI developments from the conference floor in Las Vegas.
Wang’s take: AWS has successfully advanced its enterprise and developer-focused approach to generative AI by assembling an integrated full stack of AI models and tools, all powered by AWS hardware. Businesses can realize significant cost savings and performance advantages with a full stack from AWS, Wang said. However, they can still access the most powerful Nvidia AI chips and the best generative AI models from major vendors — including AWS competitors — on AWS platforms.
“What they’re doing with the ability to give you those prebuilt models, that’s what customers want at the end of the day,” Wang said on the podcast. “They want as much prebuilt as they can get. Then they want to go customize … these things.”
As for AI factories — fully managed AI infrastructure for customers with their own AI data center and power capacity — Wang said that’s an overdue move by AWS and will help countries and regions worldwide that are building up their own sovereign AI initiatives.
“AI factories are something that they needed to do earlier, and they didn’t,” Wang said. “The AI factory is going to allow them actually to address the sovereign AI issue: ‘How do we actually get these into different countries?”