DDN has launched xFusionAI, a new Artificial Intelligence (AI) infrastructure designed to integrate training and inference capabilities into a single platform. This solution targets enterprises and research institutions, offering enhanced performance and cost efficiency for AI models ranging from lightweight architectures to those with multi-trillion parameters.

xFusionAI combines DDN's EXAScaler parallel file system with Infinia’s AI-native scalability and elasticity. The hybrid platform aims to optimize AI workflows, enabling users to train and deploy models efficiently.

The solution is engineered to support the increasing complexity and scale of AI models, boasting a 100x acceleration in large language model (LLM) training, a 10x improvement in retrieval-augmented generation performance, and a 60% reduction in inference costs.

Organizations utilizing xFusionAI are reporting substantial gains, such as 15x faster AI data center workflows, facilitating more effective model training and deployment. The platform supports diverse environments, including on-premises, cloud, and isolated systems.

According to DDN CTO Sven Oehme, “xFusionAI is the convergence of AI’s past, present, and future.” He highlights the combination of EXAScaler’s performance with Infinia’s scalability as a significant advancement in AI infrastructure.

xFusionAI aims to streamline processes and eliminate bottlenecks in the training and inference stages, providing a comprehensive solution for modern AI projects.