NVIDIA DSX Air Integrates KAI Inference Builder to Optimize AI Infrastructure

Keysight AI Inference Builder helps AI infrastructure teams validate, benchmark, and optimize inference-optimized infrastructure under real-world workload conditions. As organizations move from training large language models to deploying them at scale, inference performance has become a critical factor for capacity planning, infrastructure efficiency, and return on investment.

AI infrastructure is entering a new phase. As the industry shifts from training large language models to deploying them, organizations must optimize inference performance across increasingly complex data center environments. Inference workloads are dynamic, latency sensitive, and shaped by real user behavior, making them difficult to validate with conventional testing methods.

Keysight demonstrated Keysight AI Inference Builder operating within NVIDIA DSX Air AI factory simulation environments to model and optimize AI data center infrastructure, architectures, and performance before physical equipment is deployed. The demonstration generated realistic inference workloads throughout NVIDIA’s data center simulation environment, allowing operators to validate inference infrastructure before deploying physical equipment.

Keysight AI Inference Builder gives AI infrastructure teams a scalable way to measure, validate, and optimize real-world inference performance. The platform provides visibility into inference performance across the full stack, helping customers validate and optimize deployments before hardware reaches the rack.

The integration with NVIDIA DSX Air extends this validation into a data center simulation environment. This enables realistic inference workloads to be generated in a simulated AI factory environment so operators can validate infrastructure before physical deployment.