Intel® VTune™ Profiler

Cookbook

Download PDF
ID 766316
Date 3/22/2024
Public
Document Table of Contents
Document Table of Contents x
Cookbook
Cookbook x
Methodologies Configuration Recipes Tuning Recipes Notices and Disclaimers
Methodologies x
Analyzing Uncore Perfmon Events for Use of Intel® Data Direct I/O Technology in Intel® Xeon® Processors (NEW) Top-down Microarchitecture Analysis Method OpenMP* Code Analysis Method Custom Data Collection for Performance Analysis (NEW) 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 High Bandwidth Memory Performance on Intel® Xeon® CPU Max Series (NEW) Profiling Windows* Applications for Hybrid CPU Platforms (NEW) Viewing Analysis Results on a Web Browser (NEW) 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 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 Ingredients Prepare Your Application for Analysis Run Hotspots Analysis Identify Hot Spots in the Managed Code Optimize the Code with Loop Interchange Verify the Optimization See Also Profiling Applications in Amazon Web Services* (AWS) EC2 Instances 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 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 an Apptainer* 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 OS Thread Migration OpenMP* Imbalance and Scheduling Overhead Processor Cores Underutilization: OpenMP* Serial Time Scheduling Overhead in an Intel® oneAPI Threading Building Blocks Application PMDK Application Overhead
Cookbook

Profiling a .NET* Core Application

Profile .NET Core dynamic code with Intel® VTune™ Profiler. Locate performance hot spots in your code and optimize as needed.

Ingredients

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

  • Application: A sample C# application that adds all the elements of an integer List. The application in this recipe is used for demonstration only and it is not available for download.

  • Tools:

  • Operating system: Microsoft* Windows* 11

  • CPU: Intel microarchitecture code named Alder Lake

Prepare Your Application for Analysis

  1. Open a new command window for the .NET environment variables to take effect. Make sure that you have installed .NET Core 7.0: 

    dotnet --version

  2. Create a new listadd directory for the application: 

    mkdir C:\listadd
> cd C:\listadd

  3. Enter dotnet new console to create a new skeleton project with the following structure:

  4. Replace the contents of Program.cs in the listadd folder with C# code that adds the elements of an integer List: 

    using System;
using System.Linq;
using System.Collections.Generic;

namespace listadd
{
    class Program
    {
        static void Main(string[] args)
        {
            Console.WriteLine("Starting calculation...");            
            List<int> numbers = Enumerable.Range(1,10000).ToList();
            for (int i =0; i < 100000; i ++)
            {
                ListAdd(numbers);
            }
            
            Console.WriteLine("Calculation complete");            
        }

        static int ListAdd(List<int> candidateList)
        {
            int result = 0;
            foreach (int item in candidateList)
            {
                result += item;
            }
            
            return result;
        }        
    }
}

  5. Create listadd.dll in the C:\listadd\bin\x64\Release\net7.0 folder: 

    dotnet build -c Release

  6. Run the sample application: 

    dotnet C:\listadd\bin\x64\Release\net7.0\listadd.dll

Run Hotspots Analysis

  1. Run Intel VTune Profiler with administrator privileges.

  2. Click the New Project button on the toolbar. Specify a name for the new project, for example: dotnet.

  3. In the Analysis Target window, select local host and Launch Application target type from the left pane.

  4. In the Launch Application pane, specify the application to analyze:

    • Application: C:\Program Files\dotnet\dotnet.exe

    • Application parameters: C:\listadd\bin\x64\Release\net7.0\listadd.dll

    NOTE:

    The location of dotnet.exe depends on your environment. Use the where dotnet command to find this location.

  5. In the How pane, select Hotspots analysis. Make sure to enable Hardware Event-Based Sampling and Collect Stacks options.

  6. Click Start to run the analysis.

Identify Hot Spots in the Managed Code

When the collected analysis result opens, switch to the Bottom-up tab. Set the data grouping level to Process/Module/Function/Thread/Call Stack:

Expand dotnet.exe > listadd.dll and notice that the listadd::Program::ListAdd function occupied most of the CPU Time:

Doubleclick this hot spot function to open the source view. To view the source and disassembly code side by side, click the Assembly toggle button on the toolbar:

Use the statistics per source line/assembly instruction to identify the most time-consuming code snippets (line 24 in the example above) and work on optimizations.

Optimize the Code with Loop Interchange

Intel VTune Profiler highlights the following code line as performance-critical:

foreach (int item in candidateList)

For optimization, consider using the for loop statement. Replace the contents of Program.cs with this C# code: 

using System;
using System.Linq;
using System.Collections.Generic;

namespace listadd
{
    class Program
    {
        static void Main(string[] args)
        {
            Console.WriteLine("Starting calculation...");            
            List<int> numbers = Enumerable.Range(1,10000).ToList();
            for (int i =0; i < 100000; i ++)
            {
                ListAdd(numbers);
            }
            
            Console.WriteLine("Calculation complete");            
        }

        static int ListAdd(List<int> candidateList)
        {
            int result = 0;
            for (int i = 0; i < candidateList.Count; i++)
            {
                result += candidateList[i];
            }

            return result;
        }        
    }
}

Verify the Optimization

To verify the optimization for the updated code, repeat the Hotspots analysis.

Prior to optimization, the sample application took 511.004 ms of CPU time:

After optimization, the application ran for 430.477s, which is a 16% reduction in time over the original:

Discuss this recipe in the Analyzers community forum

NOTE:
Parent topic: Configuration Recipes
Level Two Title