NVIDIA Introduces DSX AI Factory Platform to Address Power Bottlenecks
According to NVIDIA. NVIDIA unveils its new DSX platform designed to synchronize grid, facility, and compute infrastructure for AI factories.
- • The NVIDIA DSX platform integrates grid and compute systems to improve infrastructure efficiency.
- • Strategic partnerships with energy leaders aim to create grid-aware AI factory designs.
- • New simulation and management tools help optimize power allocation and increase compute output.
Addressing Infrastructure Constraints
NVIDIA states that power is currently the primary bottleneck in AI infrastructure, with significant project backlogs in U.S. interconnection queues. The DSX platform aims to resolve this by allowing engineers to simulate the grid, substation, and AI factory as a unified system to prevent delays and stranded capacity.
Industry Collaboration and Integration
The company is working with partners including GE Vernova, Hitachi Energy, Schneider Electric, and Siemens Energy to develop reference architectures. These collaborations focus on creating resilient, scalable, and grid-aware power infrastructure that aligns with the requirements of the NVIDIA Vera Rubin AI factory design.
Operational Optimization Tools
NVIDIA DSX includes tools like DSX Flex and DSX Exchange, which allow AI factories to sync demand with grid conditions. According to NVIDIA, companies like Phaidra have utilized these integrations to optimize cooling and power allocation, reportedly delivering approximately 10% more compute within existing power envelopes.
- 1. blogs.nvidia.comOpen source