Quantcast
Channel: C#
Viewing all articles
Browse latest Browse all 1853

GPU Usage

$
0
0

For applications using Processor Graphics, configure the Intel® VTune™ Amplifier to monitor, analyze, and correlate the application performance on both the CPU and GPU.

Note

  • VTune Amplifier XE provides GPU analysis on Windows* platforms only. You cannot run GPU data collection via a Remote Desktop connection. To run the GPU data collection, run the VTune Amplifier from the target computer's console or access the computer via VNC.

  • VTune Amplifier for Systems provides GPU analysis only on Android* systems running on processors with Intel HD Graphics.

To enable the GPU analysis:

  1. Click the (standalone GUI)/ (Visual Studio IDE) New Analysis button on the VTune Amplifier toolbar.

    The Analysis Type configuration window opens.

  2. From the analysis tree on the right, choose the required analysis type.

    The right pane is updated with the configuration options for the selected analysis type. For GPU analysis, the VTune Amplifier provides three options: Analyze GPU usage, Analyze Processor Graphics hardware events, and Trace OpenCL kernels on Processor Graphics.

  3. To analyze GPU task scheduling and identify whether your application is CPU or GPU bound, select the Analyze GPU usage option.
  4. For Intel HD Graphics: To monitor the Render and GPGPU engine usage, identify which parts of the engine are loaded, correlate GPU and CPU data, select a predefined event set from the Analyze Processor Graphics hardware events drop-down menu.

    VTune Amplifier provides three presets of the hardware metrics. All presets collect data about execution units (EUs) activity: EU Array Active, EU Array Stalled, EU Array Idle, Computing Threads Started, and Core Frequency.

    • Overview event set also includes metrics that track general GPU memory accesses such as Memory Read/Write Bandwidth, GPU L3 Misses, Sampler Busy, Sampler Is Bottleneck, and GPU Memory Texture Read Bandwidth. These metrics can be useful for both graphics and compute-intensive applications.

    • Compute basic (global/local memory accesses) event group also includes metrics that distinguish accessing different types of data on a GPU: Untyped Memory Read/Write Bandwidth, Typed Memory Read/Write Transactions, SLM Read/Write Bandwidth, Render/GPGPU Command Streamer Loaded, and GPU EU Array Usage. These metrics are useful for compute-intensive workloads on the GPU.

    • Compute Extended event group includes metrics targeted only for GPU analysis on the Intel processor code name Broadwell. For other systems, this preset is disabled.

  5. For OpenCL™ kernels running on Intel HD Graphics (Windows only): To know OpenCL kernels execution time, monitor performance of each kernel per GPU metrics and identify hotspot kernels, select the Trace OpenCL kernels on Processor Graphics option.

When collection and post-processing is complete and the result is open, click to the Graphics tab to see details of GPU activity, also correlated with CPU processes and threads. For GPU metrics description, hover over the column name in the grid or right-click and select the What's This Column? context menu option.

Note

To monitor general GPU busyness over time, run the VTune Amplifier as an Administrator.

Inglese

Viewing all articles
Browse latest Browse all 1853

Trending Articles