How is NVIDIA revolutionizing AI networking infrastructure?
Kevin Deierling, Senior Vice President, Networking from NVIDIA explains:
- NVIDIA NVLink enables scale-up capabilities, allowing multiple GPUs to function as one for easier programming and enhanced tensor/pipeline parallelism
- InfiniBand serves as the gold standard for AI scale-out networking, offering 30% better performance than traditional Ethernet for highly correlated AI workloads
- Purpose-built Spectrum-X platform delivers optimized Ethernet performance for AI workloads, featuring adaptive routing and congestion control for 100,000+ GPU deployments
Check out full showcase at: https://ngi.fyi/25DCNetworkAIyt to learn more about data center networking for AI and cloud workloads
Want to be involved our video series? Contact [email protected]
