“We’re really excited about what customers can accomplish with this new infrastructure,” said Matt Garman, AWS CEO, discussing the company’s ambitious AI advancements in a recent interview. At the forefront is Project Rainier, a new chip cluster initiative powered by Amazon’s custom Trainium2 semiconductors.

The ultra clusters, featuring hundreds of thousands of chips, represent a monumental leap in computational capacity. Garman explained: “Our Trainium2 chips deliver 83 petaflops of performance from a single node, a feat enabled by Amazon-designed custom silicon and high-performance networking.” This infrastructure will quintupling the processing power available to Anthropic, a key partner, and pave the way for significantly larger and more capable AI models.

AWS’s commitment to innovation extends beyond hardware. The company recently unveiled Nova, a large language model designed for efficiency and versatility.

“Nova is quite good at executing generative workflows, pulling knowledge from indexes, and delivering very low latency at a low cost,” Garman noted. These attributes make it an attractive option for businesses seeking scalable AI solutions without breaking the bank.

While AWS develops its proprietary solutions, its partnerships remain integral. Garman is clear about the collaborative approach: “We think of this as a supplement to NVIDIA’s offerings. Customers want choice, and Trainium2 provides a 30–40% cost-performance benefit for many workloads while still supporting NVIDIA technologies.” This balanced strategy ensures AWS continues to cater to diverse client needs.

Cost efficiency is only part of the equation. Garman emphasizes customer-centricity: “Bedrock, our platform for building production AI applications, has grown fivefold in users over the past year. Customers aren’t just experimenting — they’re deeply integrating AI into their operations.” This trend reflects the increasing reliance on AI-driven insights for competitive advantage.

Garman remains optimistic about the future of AI and AWS’s role in shaping it.

“The explosion of generative AI usage creates opportunities for multiple players to succeed. It’s not about competing for slices of the pie — it’s about making the whole pie bigger,” he said. With groundbreaking projects like Rainier and Nova, AWS is poised to lead in this rapidly evolving landscape.