Post by : Badri Ariffin

OpenAI has formalised a major cloud agreement with Amazon Web Services (AWS), committing roughly $38 billion over multiple years to access AWS's high-performance infrastructure and ramp up its AI operations.

The arrangement follows OpenAI's recent organisational changes that moved it away from a non-profit structure and paved the way for broader commercial expansion. The pact also diversifies OpenAI's cloud footprint beyond its long-standing Microsoft relationship.

Across a seven-year term, OpenAI will deploy AWS EC2 UltraServers populated with thousands of Nvidia GB200 and GB300 GPUs to handle both inference for current ChatGPT models and the training of next-generation systems. The setup is engineered to scale across millions of CPU cores, with AWS highlighting gains in performance, efficiency and security; full rollout is targeted by the end of 2026, with possible extensions into 2027.

OpenAI has been widening its cloud partnerships: its foundation models are now available via Amazon Bedrock, and it maintains separate compute arrangements with Microsoft Azure, Google Cloud and Oracle to secure additional capacity for growing AI needs.

The deal marks a strategic shift for OpenAI as it seeks to accelerate research and commercial deployment without depending on a single cloud provider, underscoring the intense competition among AI players to lock in large-scale, high-performance compute resources for future innovation.

#News #Tech #AI,