NVIDIA DGX™ Cloud Benchmarking is a suite of tools, recipes, and services that take the guesswork out of measuring the performance of AI workloads and infrastructure. As an extension of NVIDIA DGX Cloud, DGX Cloud Benchmarking provides a standardized and objective means of gauging performance across platforms, essential for optimizing AI workloads and speeding outcomes.

Features

• Performance Recipes: Ready-to-use templates for evaluating the performance of specific AI workloads across hardware and software combinations.
• Performance Explorer: Identify the ideal GPU count that minimizes both total training time and costs, optimizing resources for a given workload.
• Expert Guidance: Access to NVIDIA advisors to benchmark and optimize AI workloads through the DGX Cloud Benchmarking Program.
• NVIDIA Exemplar Clouds: Initiative that raises the bar across security, usability, performance, and resiliency, matching the DGX Cloud reference architecture.

Benefits

• Improve AI Workload Performance: Unlock the full potential of AI infrastructure with DGX Cloud Benchmarking, optimizing AI workload environments.
• Measure True AI Workload TCO: Determine the fastest time to train or desired GPU scale and at what cost using real-time and end-to-end performance data.
• Accelerate AI Outcomes: Tune and optimize AI workloads according to end-to-end metrics tailored to the performance of modern generative AI applications.
• Benchmark Beyond the FLOPS: Evaluate infrastructure software, cloud platforms, and application configurations to gain a holistic view of workload performance.
• Standardized Performance Gauging: Obtain a standardized and objective means of gauging platform performance, understanding expected performance for given workloads or use cases.