The NVIDIA® Nsight™ Perf SDK is a graphics profiling toolbox for DirectX, Vulkan, and OpenGL, enabling developers to collect GPU performance metrics directly from their applications. With just a few lines of code, developers can set up GPU performance metrics collection, enabling real-time performance triage and detailed profiling.

Features

• Realtime Perf Triage: Enable high-level performance triage via real-time collection and on-screen visualization of GPU performance metrics. The GPU Periodic Sampler collects device-level metrics at high sampling rates with low overhead.
• Profile In-Application: Integrate GPU performance metric collection into your application or graphics developer tool of choice. Activate profiling from custom programmatic triggers, choose the list of GPU metrics to collect, and customize your output.
• Upgrade Your CI/CD: Generate detailed profiler reports on every developer and artist change. Add dedicated performance regression criteria by inspecting GPU metric values.
• Realtime Performance HUD: Add continuous performance metrics collection to your code, leveraging the built-in HUD renderer for real-time, high-level performance triage. Explore panels with metrics on SM, L2 cache, ROP, VRAM, and other subunits.
• Timeline Viewer: Examine a snapshot of your application’s performance with one-shot sampling mode, visualizing hardware activity with minimal overhead.
• HTML Profiler Report Generator: Generate detailed profiler reports with minimal effort, providing a top-down representation of GPU performance and fast navigation to performance limiters.

Benefits

• Enhanced Performance: Optimize graphics applications by identifying and resolving performance bottlenecks.
• Detailed Insights: Gain comprehensive insights into GPU performance, from memory workload to source code efficiency.
• Flexibility: Supports various development environments and platforms, including desktop, mobile, cloud, and HPC.
• Ease of Use: Offers user-friendly interfaces and integration with popular development environments.
• Advanced Profiling: Enables detailed performance analysis and debugging for CUDA, graphics, and deep learning applications.