Intel VTune Profiler Performance Analysis Cookbook

Download PDF
ID 766316
Date 5/19/2023
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents
Document Table of Contents x
Intel® VTune™ Profiler Performance Analysis Cookbook
Intel® VTune™ Profiler Performance Analysis Cookbook x
Methodologies Configuration Recipes Tuning Recipes Notices and Disclaimers
Methodologies x
Top-down Microarchitecture Analysis Method OpenMP* Code Analysis Method Software Optimization for Intel® GPUs (NEW) Core Utilization in DPDK Apps PCIe Traffic in DPDK Apps DPDK Event Device Profiling Effective Utilization of Intel® Data Direct I/O Technology Compile a Portable Optimized Binary with the Latest Instruction Set
Configuration Recipes x
Profiling Machine Learning Applications (NEW) Profiling Single-Node Kubernetes* Applications (NEW) Analyzing Hot Code Paths Using Flame Graphs (NEW) Improving Hotspot Observability in a C++ Application Using Flame Graphs Measuring Performance Impact of NUMA in Multi-Processor Systems (NEW) Profiling Games built with Unity* (NEW) Profiling Games built with Unreal Engine* (NEW) Profiling Java Applications as a Remote User (NEW) Profiling JavaScript* Code in Node.js* Analyzing CPU and FPGA (Intel® Arria® 10 GX) Interaction Profiling a .NET* Core Application Profiling Applications in Amazon Web Services* (AWS) EC2 Instances Ingredients Create and Configure a Virtual Machine Instance Connect to the EC 2 Instance Configure the Instance for Profiling Run Hotspots Analysis Enabling Performance Profiling in GitLab* CI Configuring a Hyper-V* Virtual Machine for Hardware-Based Hotspots Analysis Profiling an Application for Performance Anomalies (NEW) Profiling an OpenMP* Offload Application running on a GPU (NEW) Profiling a SYCL* Application running on a GPU Profiling an FPGA-driven SYCL* Application Profiling Hardware Without Intel Sampling Drivers Profiling MPI Applications Profiling Docker* Containers Profiling a Remote Target Through a Proxy Server (NEW) Profiling in a Singularity* Container Profiling Linux*, Android*, and QNX* System Boot Time Using Intel® VTune™ Profiler Server with Visual Studio Code and Intel® DevCloud for oneAPI (NEW) Using Intel® VTune™ Profiler Server in HPC Clusters Using the Command-Line Interface to Analyze the Performance of a SYCL* Application running on a GPU (NEW)
Tuning Recipes x
Cache-Related Latency Issues in Segmented Cache Environment False Sharing Frequent DRAM Accesses Poor Port Utilization Page Faults Instruction Cache Misses Inefficient Synchronization Inefficient TCP/IP Synchronization OS Thread Migration OpenMP* Imbalance and Scheduling Overhead Processor Cores Underutilization: OpenMP* Serial Time Scheduling Overhead in Intel® Threading Building Blocks (Intel® TBB) Apps PMDK Application Overhead
Intel® VTune™ Profiler Performance Analysis Cookbook

Profiling Applications in Amazon Web Services* (AWS) EC2 Instances

This recipe helps you set up a VM instance in AWS for performance profiling with Intel® VTune™ Profiler.

Ingredients

This section lists the hardware and software tools used for the performance analysis scenario.

  • Application:matrix. The application is used as a demo and not available for download.

  • Tools: Intel® VTune™ Profiler 2020 - Hotspots analysis

    NOTE:

    • Starting with the 2020 release, Intel® VTune™ Amplifier has been renamed to Intel® VTune™ Profiler.

    • Most recipes in the Intel® VTune™ Profiler Performance Analysis Cookbook are flexible. You can apply them to different versions of Intel® VTune™ Profiler. In some cases, minor adjustments may be required.

    • Get the latest version of Intel® VTune™ Profiler:

Create and Configure a Virtual Machine Instance

  1. On the Amazon Web Services site, click Sign In to the Console to log in to your account.

    If you do not have an account, create it.

  2. Enter the Amazon EC2 Dashboard, select Compute > EC2.

  3. Click Launch Instance to create and configure your virtual machine with a wizard.

    In the configuration wizard, you can specify an OS and specification of the instance that fits your case (number of CPUs, memory, storage, and networking capacity).

  4. In Step 1: Choose an Amazon Machine Image (AMI), select Ubuntu Server 18.04 LTSlick.

  5. In Step 2: Choose an Instance Type, select an instance and click Next: Configure Instance Details.

    NOTE:

    Recommended system requirements for the VTune Profiler is 4 GB of RAM and at least 10Gb free disk space.

  6. In Step 3: Configure Instance Details, enable Auto-assign Public IP and click Next: Add Storage.

  7. In Step 4: Add Storage, increase the size of the storage device and click Next: Add Tags.

  8. [Optional] In Step 5: Add Tags, add a tag (for example, VTune Demo), then click Review and Launch.

  9. In Step 7: Review Instance Launch, click Launch.

  10. Create a key pair:

    1. Select Create a new key pair and assign a name (for example, MyEC2Instance).

    2. Click Download Key Pair to download the key.

    3. Select Launch Instances to complete the setup.

      The instance is launched.

  11. Click View Instances.

    When the Instance State is updated to running, proceed to the next step.

Connect to the EC 2 Instance

Select the EC2 instance you created and click the Connect tab to connect to your instance via SSH.

PuTTY* does not natively support the private key format (.pem) generated by Amazon EC2. You should convert your private key (.pem) to the required PuTTY format (.ppk) using PuTTYgen before attempting to connect to your instance using PuTTY:

  1. Convert your private key:

    1. Start PuTTYgen.

    2. For the Type of key to generate, select SSH-1 (RSA).

    3. Click Load and select your .pem file.

    4. Click Save private key to save the key in the format that PuTTY can use.

  2. Start PuTTY and specify a Public DNS for the instance and your private key (.ppk) file in the Connection > SSH > Auth pane.

  3. Save this configuration and click Open.

    You are logged in to the instance.

NOTE:

You need to install the Xubuntu* desktop on your running instance and setup a VNC* Server if you want to access the system via VNC.

Configure the Instance for Profiling

Prepare the target instance for profiling by setting /proc/sys/kernel/yama/ptrace_scope to 0:

echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope

Run Hotspots Analysis

Run the Hotspots analysis using any of the following methods:

Run a remote collection from the locally installed VTune Profiler using SSH:

NOTE:

Remote hardware event-based sampling analysis is not available in the AWS EC2 virtualized environment with an exception for the EC2 bare metal instances (i3.metal instance family) that allow direct access to bare metal infrastructure and full profiling capability of the VTune Profiler.

  1. Create a project in the VTune Profiler.

    The Configure Analysis window opens.

  2. In the WHERE pane, select Remote Linux (SSH) and enter ubuntu@<name of putty configuration> for the SSH destination.

    At this point, VTune Profiler attempts to connect to the remote system and install the binaries necessary for collection to the specified folder.

  3. In the WHAT pane, specify the location of your application and its working directory.

  4. In the HOW pane, select the User-Mode Sampling collection mode for the Hotspots analysis (open by default).

  5. Click to start the collection.

    The results are automatically copied to your local system for analysis.

Run VTune Profiler directly from the AWS instance:

  1. Install the VTune Profiler as described in the Installation Guide.

  2. Run the VTune Profiler. For example:

    sudo <vtune_install_dir>/vtune_profiler/bin64/vtune-gui

  3. Create a project.

    The Configure Analysis window opens.

  4. In the WHERE pane, select Local Host.

  5. In the WHAT pane, specify the location of your application and its working directory.

  6. In the HOW pane, select a preferred collection mode for the Hotspots analysis, for example: Hardware Event-Based Sampling.

  7. Click to start the collection.

    The analysis result opens in the default Hotspots by CPU Utilization viewpoint.

Parent topic: Configuration Recipes
Level Two Title