Nvidia and Microsoft collaborated to develop a new specialized computer architecture. This hardware integration optimizes the synergy between GPU clusters and cloud infrastructure. The move targets massive scale-out capabilities for training next-generation frontier models. It provides developers with higher throughput and reduced latency for complex AI workloads.