The HPE Cray XD670 is built for AI-intensive workloads like LLM training and NLP, offering unmatched performance and scalability. Powered by up to eight NVIDIA H200 or H100 GPUs, it delivers top benchmark results in MLPerf Inference and Training. The system supports high memory bandwidth and energy-efficient performance, ideal for enterprises scaling AI beyond pilots. With sustainable cooling options—including direct liquid cooling—and the ability to scale from one to thousands of nodes, XD670 meets growing AI demands. Backed by HPE Services and Complete Care, it combines expert support with flexible financing for long-term impact and sustainability.

Join for free to read