Oracle and AMD have unveiled a major collaboration to equip Oracle Cloud Infrastructure (OCI) with AMD Instinct MI355X GPUs, enabling zettascale AI clustersāon-demand capabilities that can scale up to 131,072 GPUs. This marks a significant leap in cloud compute power, offering customers over twice the price-performance for large-scale AI training and inference compared to previous setups.
The move signals Oracleās aggressive push to compete with hyperscale providers like AWS, Azure, and Google Cloud in the high-performance AI space. Businesses developing advanced AI modelsāespecially in generative or agentic AIācan now access supercomputing-class resources on OCI. This opens new doors for AI startups, research labs, and enterprises in India and worldwide to train large language and vision models without building their own expensive on-prem infrastructure.
