mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2026-10-23 21:28:17 +08:00
add and update samples for CUDA 11.6
This commit is contained in:
Rutwik Choughule
18
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/c_cpp_properties.json
vendored
Normal file
18
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/c_cpp_properties.json
vendored
Normal file
@@ -0,0 +1,18 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "Linux",
|
||||
"includePath": [
|
||||
"${workspaceFolder}/**",
|
||||
"${workspaceFolder}/../../../Common"
|
||||
],
|
||||
"defines": [],
|
||||
"compilerPath": "/usr/local/cuda/bin/nvcc",
|
||||
"cStandard": "gnu17",
|
||||
"cppStandard": "gnu++14",
|
||||
"intelliSenseMode": "linux-gcc-x64",
|
||||
"configurationProvider": "ms-vscode.makefile-tools"
|
||||
}
|
||||
],
|
||||
"version": 4
|
||||
}
|
||||
7
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/extensions.json
vendored
Normal file
7
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/extensions.json
vendored
Normal file
@@ -0,0 +1,7 @@
|
||||
{
|
||||
"recommendations": [
|
||||
"nvidia.nsight-vscode-edition",
|
||||
"ms-vscode.cpptools",
|
||||
"ms-vscode.makefile-tools"
|
||||
]
|
||||
}
|
||||
10
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/launch.json
vendored
Normal file
10
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/launch.json
vendored
Normal file
@@ -0,0 +1,10 @@
|
||||
{
|
||||
"configurations": [
|
||||
{
|
||||
"name": "CUDA C++: Launch",
|
||||
"type": "cuda-gdb",
|
||||
"request": "launch",
|
||||
"program": "${workspaceFolder}/MC_SingleAsianOptionP"
|
||||
}
|
||||
]
|
||||
}
|
||||
15
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/tasks.json
vendored
Normal file
15
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/.vscode/tasks.json
vendored
Normal file
@@ -0,0 +1,15 @@
|
||||
{
|
||||
"version": "2.0.0",
|
||||
"tasks": [
|
||||
{
|
||||
"label": "sample",
|
||||
"type": "shell",
|
||||
"command": "make dbg=1",
|
||||
"problemMatcher": ["$nvcc"],
|
||||
"group": {
|
||||
"kind": "build",
|
||||
"isDefault": true
|
||||
}
|
||||
}
|
||||
]
|
||||
}
|
||||
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2017
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "MC_SingleAsianOptionP", "MC_SingleAsianOptionP_vs2017.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
@@ -0,0 +1,117 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>MC_SingleAsianOptionP_vs2017</RootNamespace>
|
||||
<ProjectName>MC_SingleAsianOptionP</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(WindowsTargetPlatformVersion)'==''">
|
||||
<LatestTargetPlatformVersion>$([Microsoft.Build.Utilities.ToolLocationHelper]::GetLatestSDKTargetPlatformVersion('Windows', '10.0'))</LatestTargetPlatformVersion>
|
||||
<WindowsTargetPlatformVersion Condition="'$(WindowsTargetPlatformVersion)' == ''">$(LatestTargetPlatformVersion)</WindowsTargetPlatformVersion>
|
||||
<TargetPlatformVersion>$(WindowsTargetPlatformVersion)</TargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v141</PlatformToolset>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../../Common;$(CudaToolkitIncludeDir);</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>curand.lib;cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/MC_SingleAsianOptionP.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" --threads 0 </AdditionalOptions>
|
||||
<Include>./;../../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="src/main.cpp" />
|
||||
<CudaCompile Include="src/pricingengine.cu" />
|
||||
<ClCompile Include="src/test.cpp" />
|
||||
<ClInclude Include="inc/asianoption.h" />
|
||||
<ClInclude Include="inc/cudasharedmem.h" />
|
||||
<ClInclude Include="inc/pricingengine.h" />
|
||||
<ClInclude Include="inc/test.h" />
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2019
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "MC_SingleAsianOptionP", "MC_SingleAsianOptionP_vs2019.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
@@ -0,0 +1,113 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>MC_SingleAsianOptionP_vs2019</RootNamespace>
|
||||
<ProjectName>MC_SingleAsianOptionP</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v142</PlatformToolset>
|
||||
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../../Common;$(CudaToolkitIncludeDir);</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>curand.lib;cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/MC_SingleAsianOptionP.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" --threads 0 </AdditionalOptions>
|
||||
<Include>./;../../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="src/main.cpp" />
|
||||
<CudaCompile Include="src/pricingengine.cu" />
|
||||
<ClCompile Include="src/test.cpp" />
|
||||
<ClInclude Include="inc/asianoption.h" />
|
||||
<ClInclude Include="inc/cudasharedmem.h" />
|
||||
<ClInclude Include="inc/pricingengine.h" />
|
||||
<ClInclude Include="inc/test.h" />
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
@@ -0,0 +1,20 @@
|
||||
|
||||
Microsoft Visual Studio Solution File, Format Version 12.00
|
||||
# Visual Studio 2022
|
||||
Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "MC_SingleAsianOptionP", "MC_SingleAsianOptionP_vs2022.vcxproj", "{997E0757-EA74-4A4E-A0FC-47D8C8831A15}"
|
||||
EndProject
|
||||
Global
|
||||
GlobalSection(SolutionConfigurationPlatforms) = preSolution
|
||||
Debug|x64 = Debug|x64
|
||||
Release|x64 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(ProjectConfigurationPlatforms) = postSolution
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.ActiveCfg = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Debug|x64.Build.0 = Debug|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.ActiveCfg = Release|x64
|
||||
{997E0757-EA74-4A4E-A0FC-47D8C8831A15}.Release|x64.Build.0 = Release|x64
|
||||
EndGlobalSection
|
||||
GlobalSection(SolutionProperties) = preSolution
|
||||
HideSolutionNode = FALSE
|
||||
EndGlobalSection
|
||||
EndGlobal
|
||||
@@ -0,0 +1,113 @@
|
||||
<?xml version="1.0" encoding="utf-8"?>
|
||||
<Project DefaultTargets="Build" ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
|
||||
<PropertyGroup>
|
||||
<CUDAPropsPath Condition="'$(CUDAPropsPath)'==''">$(VCTargetsPath)\BuildCustomizations</CUDAPropsPath>
|
||||
</PropertyGroup>
|
||||
<ItemGroup Label="ProjectConfigurations">
|
||||
<ProjectConfiguration Include="Debug|x64">
|
||||
<Configuration>Debug</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
<ProjectConfiguration Include="Release|x64">
|
||||
<Configuration>Release</Configuration>
|
||||
<Platform>x64</Platform>
|
||||
</ProjectConfiguration>
|
||||
</ItemGroup>
|
||||
<PropertyGroup Label="Globals">
|
||||
<ProjectGuid>{997E0757-EA74-4A4E-A0FC-47D8C8831A15}</ProjectGuid>
|
||||
<RootNamespace>MC_SingleAsianOptionP_vs2022</RootNamespace>
|
||||
<ProjectName>MC_SingleAsianOptionP</ProjectName>
|
||||
<CudaToolkitCustomDir />
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
|
||||
<PropertyGroup>
|
||||
<ConfigurationType>Application</ConfigurationType>
|
||||
<CharacterSet>MultiByte</CharacterSet>
|
||||
<PlatformToolset>v143</PlatformToolset>
|
||||
<WindowsTargetPlatformVersion>10.0</WindowsTargetPlatformVersion>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<UseDebugLibraries>true</UseDebugLibraries>
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Configuration)'=='Release'">
|
||||
<WholeProgramOptimization>true</WholeProgramOptimization>
|
||||
</PropertyGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
|
||||
<ImportGroup Label="ExtensionSettings">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.props" />
|
||||
</ImportGroup>
|
||||
<ImportGroup Label="PropertySheets">
|
||||
<Import Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" />
|
||||
</ImportGroup>
|
||||
<PropertyGroup Label="UserMacros" />
|
||||
<PropertyGroup>
|
||||
<IntDir>$(Platform)/$(Configuration)/</IntDir>
|
||||
<IncludePath>$(IncludePath)</IncludePath>
|
||||
<CodeAnalysisRuleSet>AllRules.ruleset</CodeAnalysisRuleSet>
|
||||
<CodeAnalysisRules />
|
||||
<CodeAnalysisRuleAssemblies />
|
||||
</PropertyGroup>
|
||||
<PropertyGroup Condition="'$(Platform)'=='x64'">
|
||||
<OutDir>../../../bin/win64/$(Configuration)/</OutDir>
|
||||
</PropertyGroup>
|
||||
<ItemDefinitionGroup>
|
||||
<ClCompile>
|
||||
<WarningLevel>Level3</WarningLevel>
|
||||
<PreprocessorDefinitions>WIN32;_MBCS;%(PreprocessorDefinitions)</PreprocessorDefinitions>
|
||||
<AdditionalIncludeDirectories>./;$(CudaToolkitDir)/include;../../../Common;$(CudaToolkitIncludeDir);</AdditionalIncludeDirectories>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<SubSystem>Console</SubSystem>
|
||||
<AdditionalDependencies>curand.lib;cudart_static.lib;kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies)</AdditionalDependencies>
|
||||
<AdditionalLibraryDirectories>$(CudaToolkitLibDir);</AdditionalLibraryDirectories>
|
||||
<OutputFile>$(OutDir)/MC_SingleAsianOptionP.exe</OutputFile>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<CodeGeneration>compute_35,sm_35;compute_37,sm_37;compute_50,sm_50;compute_52,sm_52;compute_60,sm_60;compute_61,sm_61;compute_70,sm_70;compute_75,sm_75;compute_80,sm_80;compute_86,sm_86;</CodeGeneration>
|
||||
<AdditionalOptions>-Xcompiler "/wd 4819" --threads 0 </AdditionalOptions>
|
||||
<Include>./;../../../Common</Include>
|
||||
<Defines>WIN32</Defines>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Debug'">
|
||||
<ClCompile>
|
||||
<Optimization>Disabled</Optimization>
|
||||
<RuntimeLibrary>MultiThreadedDebug</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>true</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>Default</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MTd</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemDefinitionGroup Condition="'$(Configuration)'=='Release'">
|
||||
<ClCompile>
|
||||
<Optimization>MaxSpeed</Optimization>
|
||||
<RuntimeLibrary>MultiThreaded</RuntimeLibrary>
|
||||
</ClCompile>
|
||||
<Link>
|
||||
<GenerateDebugInformation>false</GenerateDebugInformation>
|
||||
<LinkTimeCodeGeneration>UseLinkTimeCodeGeneration</LinkTimeCodeGeneration>
|
||||
</Link>
|
||||
<CudaCompile>
|
||||
<Runtime>MT</Runtime>
|
||||
<TargetMachinePlatform>64</TargetMachinePlatform>
|
||||
</CudaCompile>
|
||||
</ItemDefinitionGroup>
|
||||
<ItemGroup>
|
||||
<ClCompile Include="src/main.cpp" />
|
||||
<CudaCompile Include="src/pricingengine.cu" />
|
||||
<ClCompile Include="src/test.cpp" />
|
||||
<ClInclude Include="inc/asianoption.h" />
|
||||
<ClInclude Include="inc/cudasharedmem.h" />
|
||||
<ClInclude Include="inc/pricingengine.h" />
|
||||
<ClInclude Include="inc/test.h" />
|
||||
</ItemGroup>
|
||||
<Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
|
||||
<ImportGroup Label="ExtensionTargets">
|
||||
<Import Project="$(CUDAPropsPath)\CUDA 11.6.targets" />
|
||||
</ImportGroup>
|
||||
</Project>
|
||||
348
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/Makefile
Normal file
348
Samples/2_Concepts_and_Techniques/MC_SingleAsianOptionP/Makefile
Normal file
@@ -0,0 +1,348 @@
|
||||
################################################################################
|
||||
# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
#
|
||||
# Redistribution and use in source and binary forms, with or without
|
||||
# modification, are permitted provided that the following conditions
|
||||
# are met:
|
||||
# * Redistributions of source code must retain the above copyright
|
||||
# notice, this list of conditions and the following disclaimer.
|
||||
# * Redistributions in binary form must reproduce the above copyright
|
||||
# notice, this list of conditions and the following disclaimer in the
|
||||
# documentation and/or other materials provided with the distribution.
|
||||
# * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
# contributors may be used to endorse or promote products derived
|
||||
# from this software without specific prior written permission.
|
||||
#
|
||||
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
#
|
||||
################################################################################
|
||||
#
|
||||
# Makefile project only supported on Mac OS X and Linux Platforms)
|
||||
#
|
||||
################################################################################
|
||||
|
||||
# Location of the CUDA Toolkit
|
||||
CUDA_PATH ?= /usr/local/cuda
|
||||
|
||||
##############################
|
||||
# start deprecated interface #
|
||||
##############################
|
||||
ifeq ($(x86_64),1)
|
||||
$(info WARNING - x86_64 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=x86_64 instead)
|
||||
TARGET_ARCH ?= x86_64
|
||||
endif
|
||||
ifeq ($(ARMv7),1)
|
||||
$(info WARNING - ARMv7 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=armv7l instead)
|
||||
TARGET_ARCH ?= armv7l
|
||||
endif
|
||||
ifeq ($(aarch64),1)
|
||||
$(info WARNING - aarch64 variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=aarch64 instead)
|
||||
TARGET_ARCH ?= aarch64
|
||||
endif
|
||||
ifeq ($(ppc64le),1)
|
||||
$(info WARNING - ppc64le variable has been deprecated)
|
||||
$(info WARNING - please use TARGET_ARCH=ppc64le instead)
|
||||
TARGET_ARCH ?= ppc64le
|
||||
endif
|
||||
ifneq ($(GCC),)
|
||||
$(info WARNING - GCC variable has been deprecated)
|
||||
$(info WARNING - please use HOST_COMPILER=$(GCC) instead)
|
||||
HOST_COMPILER ?= $(GCC)
|
||||
endif
|
||||
ifneq ($(abi),)
|
||||
$(error ERROR - abi variable has been removed)
|
||||
endif
|
||||
############################
|
||||
# end deprecated interface #
|
||||
############################
|
||||
|
||||
# architecture
|
||||
HOST_ARCH := $(shell uname -m)
|
||||
TARGET_ARCH ?= $(HOST_ARCH)
|
||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le armv7l))
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifneq (,$(filter $(TARGET_ARCH),x86_64 aarch64 sbsa ppc64le))
|
||||
TARGET_SIZE := 64
|
||||
else ifneq (,$(filter $(TARGET_ARCH),armv7l))
|
||||
TARGET_SIZE := 32
|
||||
endif
|
||||
else
|
||||
TARGET_SIZE := $(shell getconf LONG_BIT)
|
||||
endif
|
||||
else
|
||||
$(error ERROR - unsupported value $(TARGET_ARCH) for TARGET_ARCH!)
|
||||
endif
|
||||
|
||||
# sbsa and aarch64 systems look similar. Need to differentiate them at host level for now.
|
||||
ifeq ($(HOST_ARCH),aarch64)
|
||||
ifeq ($(CUDA_PATH)/targets/sbsa-linux,$(shell ls -1d $(CUDA_PATH)/targets/sbsa-linux 2>/dev/null))
|
||||
HOST_ARCH := sbsa
|
||||
TARGET_ARCH := sbsa
|
||||
endif
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq (,$(filter $(HOST_ARCH)-$(TARGET_ARCH),aarch64-armv7l x86_64-armv7l x86_64-aarch64 x86_64-sbsa x86_64-ppc64le))
|
||||
$(error ERROR - cross compiling from $(HOST_ARCH) to $(TARGET_ARCH) is not supported!)
|
||||
endif
|
||||
endif
|
||||
|
||||
# When on native aarch64 system with userspace of 32-bit, change TARGET_ARCH to armv7l
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_SIZE),aarch64-aarch64-32)
|
||||
TARGET_ARCH = armv7l
|
||||
endif
|
||||
|
||||
# operating system
|
||||
HOST_OS := $(shell uname -s 2>/dev/null | tr "[:upper:]" "[:lower:]")
|
||||
TARGET_OS ?= $(HOST_OS)
|
||||
ifeq (,$(filter $(TARGET_OS),linux darwin qnx android))
|
||||
$(error ERROR - unsupported value $(TARGET_OS) for TARGET_OS!)
|
||||
endif
|
||||
|
||||
# host compiler
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
ifeq ($(shell expr `xcodebuild -version | grep -i xcode | awk '{print $$2}' | cut -d'.' -f1` \>= 5),1)
|
||||
HOST_COMPILER ?= clang++
|
||||
endif
|
||||
else ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq ($(HOST_ARCH)-$(TARGET_ARCH),x86_64-armv7l)
|
||||
ifeq ($(TARGET_OS),linux)
|
||||
HOST_COMPILER ?= arm-linux-gnueabihf-g++
|
||||
else ifeq ($(TARGET_OS),qnx)
|
||||
ifeq ($(QNX_HOST),)
|
||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
||||
endif
|
||||
ifeq ($(QNX_TARGET),)
|
||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
||||
endif
|
||||
export QNX_HOST
|
||||
export QNX_TARGET
|
||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/arm-unknown-nto-qnx6.6.0eabi-g++
|
||||
else ifeq ($(TARGET_OS),android)
|
||||
HOST_COMPILER ?= arm-linux-androideabi-g++
|
||||
endif
|
||||
else ifeq ($(TARGET_ARCH),aarch64)
|
||||
ifeq ($(TARGET_OS), linux)
|
||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
||||
else ifeq ($(TARGET_OS),qnx)
|
||||
ifeq ($(QNX_HOST),)
|
||||
$(error ERROR - QNX_HOST must be passed to the QNX host toolchain)
|
||||
endif
|
||||
ifeq ($(QNX_TARGET),)
|
||||
$(error ERROR - QNX_TARGET must be passed to the QNX target toolchain)
|
||||
endif
|
||||
export QNX_HOST
|
||||
export QNX_TARGET
|
||||
HOST_COMPILER ?= $(QNX_HOST)/usr/bin/q++
|
||||
else ifeq ($(TARGET_OS), android)
|
||||
HOST_COMPILER ?= aarch64-linux-android-clang++
|
||||
endif
|
||||
else ifeq ($(TARGET_ARCH),sbsa)
|
||||
HOST_COMPILER ?= aarch64-linux-gnu-g++
|
||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
||||
HOST_COMPILER ?= powerpc64le-linux-gnu-g++
|
||||
endif
|
||||
endif
|
||||
HOST_COMPILER ?= g++
|
||||
NVCC := $(CUDA_PATH)/bin/nvcc -ccbin $(HOST_COMPILER)
|
||||
|
||||
# internal flags
|
||||
NVCCFLAGS := -m${TARGET_SIZE}
|
||||
CCFLAGS :=
|
||||
LDFLAGS :=
|
||||
|
||||
# build flags
|
||||
ifeq ($(TARGET_OS),darwin)
|
||||
LDFLAGS += -rpath $(CUDA_PATH)/lib
|
||||
CCFLAGS += -arch $(HOST_ARCH)
|
||||
else ifeq ($(HOST_ARCH)-$(TARGET_ARCH)-$(TARGET_OS),x86_64-armv7l-linux)
|
||||
LDFLAGS += --dynamic-linker=/lib/ld-linux-armhf.so.3
|
||||
CCFLAGS += -mfloat-abi=hard
|
||||
else ifeq ($(TARGET_OS),android)
|
||||
LDFLAGS += -pie
|
||||
CCFLAGS += -fpie -fpic -fexceptions
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_ARCH),$(HOST_ARCH))
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
||||
ifneq ($(TARGET_FS),)
|
||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
||||
endif
|
||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/arm-linux-gnueabihf
|
||||
endif
|
||||
endif
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
||||
ifneq ($(TARGET_FS),)
|
||||
GCCVERSIONLTEQ46 := $(shell expr `$(HOST_COMPILER) -dumpversion` \<= 4.6)
|
||||
ifeq ($(GCCVERSIONLTEQ46),1)
|
||||
CCFLAGS += --sysroot=$(TARGET_FS)
|
||||
endif
|
||||
LDFLAGS += --sysroot=$(TARGET_FS)
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib -L$(TARGET_FS)/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/lib/aarch64-linux-gnu -L$(TARGET_FS)/lib/aarch64-linux-gnu
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib -L$(TARGET_FS)/usr/lib
|
||||
LDFLAGS += -rpath-link=$(TARGET_FS)/usr/lib/aarch64-linux-gnu -L$(TARGET_FS)/usr/lib/aarch64-linux-gnu
|
||||
LDFLAGS += --unresolved-symbols=ignore-in-shared-libs
|
||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include -I$(TARGET_FS)/usr/include/libdrm
|
||||
CCFLAGS += -isystem=$(TARGET_FS)/usr/include/aarch64-linux-gnu -I$(TARGET_FS)/usr/include/aarch64-linux-gnu
|
||||
endif
|
||||
endif
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
||||
NVCCFLAGS += -D_QNX_SOURCE
|
||||
NVCCFLAGS += --qpp-config 8.3.0,gcc_ntoaarch64le
|
||||
CCFLAGS += -DWIN_INTERFACE_CUSTOM -I/usr/include/aarch64-qnx-gnu
|
||||
LDFLAGS += -lsocket
|
||||
LDFLAGS += -L/usr/lib/aarch64-qnx-gnu
|
||||
CCFLAGS += "-Wl\,-rpath-link\,/usr/lib/aarch64-qnx-gnu"
|
||||
ifdef TARGET_OVERRIDE
|
||||
LDFLAGS += -lslog2
|
||||
endif
|
||||
|
||||
ifneq ($(TARGET_FS),)
|
||||
LDFLAGS += -L$(TARGET_FS)/usr/lib
|
||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/lib"
|
||||
LDFLAGS += -L$(TARGET_FS)/usr/libnvidia
|
||||
CCFLAGS += "-Wl\,-rpath-link\,$(TARGET_FS)/usr/libnvidia"
|
||||
CCFLAGS += -I$(TARGET_FS)/../include
|
||||
endif
|
||||
endif
|
||||
endif
|
||||
|
||||
ifdef TARGET_OVERRIDE # cuda toolkit targets override
|
||||
NVCCFLAGS += -target-dir $(TARGET_OVERRIDE)
|
||||
endif
|
||||
|
||||
# Install directory of different arch
|
||||
CUDA_INSTALL_TARGET_DIR :=
|
||||
ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-gnueabihf/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),sbsa-linux)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/sbsa-linux/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-android)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/armv7-linux-androideabi/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-android)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-linux-androideabi/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),armv7l-qnx)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/ARMv7-linux-QNX/
|
||||
else ifeq ($(TARGET_ARCH)-$(TARGET_OS),aarch64-qnx)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/aarch64-qnx/
|
||||
else ifeq ($(TARGET_ARCH),ppc64le)
|
||||
CUDA_INSTALL_TARGET_DIR = targets/ppc64le-linux/
|
||||
endif
|
||||
|
||||
# Debug build flags
|
||||
ifeq ($(dbg),1)
|
||||
NVCCFLAGS += -g -G
|
||||
BUILD_TYPE := debug
|
||||
else
|
||||
BUILD_TYPE := release
|
||||
endif
|
||||
|
||||
ALL_CCFLAGS :=
|
||||
ALL_CCFLAGS += $(NVCCFLAGS)
|
||||
ALL_CCFLAGS += $(EXTRA_NVCCFLAGS)
|
||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(CCFLAGS))
|
||||
ALL_CCFLAGS += $(addprefix -Xcompiler ,$(EXTRA_CCFLAGS))
|
||||
|
||||
SAMPLE_ENABLED := 1
|
||||
|
||||
ALL_LDFLAGS :=
|
||||
ALL_LDFLAGS += $(ALL_CCFLAGS)
|
||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(LDFLAGS))
|
||||
ALL_LDFLAGS += $(addprefix -Xlinker ,$(EXTRA_LDFLAGS))
|
||||
|
||||
# Common includes and paths for CUDA
|
||||
INCLUDES := -I../../../Common
|
||||
LIBRARIES :=
|
||||
|
||||
################################################################################
|
||||
|
||||
# Gencode arguments
|
||||
ifeq ($(TARGET_ARCH),$(filter $(TARGET_ARCH),armv7l aarch64 sbsa))
|
||||
SMS ?= 53 61 70 72 75 80 86 87
|
||||
else
|
||||
SMS ?= 35 37 50 52 60 61 70 75 80 86
|
||||
endif
|
||||
|
||||
ifeq ($(SMS),)
|
||||
$(info >>> WARNING - no SM architectures have been specified - waiving sample <<<)
|
||||
SAMPLE_ENABLED := 0
|
||||
endif
|
||||
|
||||
ifeq ($(GENCODE_FLAGS),)
|
||||
# Generate SASS code for each SM architecture listed in $(SMS)
|
||||
$(foreach sm,$(SMS),$(eval GENCODE_FLAGS += -gencode arch=compute_$(sm),code=sm_$(sm)))
|
||||
|
||||
# Generate PTX code from the highest SM architecture in $(SMS) to guarantee forward-compatibility
|
||||
HIGHEST_SM := $(lastword $(sort $(SMS)))
|
||||
ifneq ($(HIGHEST_SM),)
|
||||
GENCODE_FLAGS += -gencode arch=compute_$(HIGHEST_SM),code=compute_$(HIGHEST_SM)
|
||||
endif
|
||||
endif
|
||||
|
||||
ALL_CCFLAGS += --threads 0 --std=c++11
|
||||
|
||||
LIBRARIES += -lcurand
|
||||
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
EXEC ?= @echo "[@]"
|
||||
endif
|
||||
|
||||
################################################################################
|
||||
|
||||
# Target rules
|
||||
all: build
|
||||
|
||||
build: MC_SingleAsianOptionP
|
||||
|
||||
check.deps:
|
||||
ifeq ($(SAMPLE_ENABLED),0)
|
||||
@echo "Sample will be waived due to the above missing dependencies"
|
||||
else
|
||||
@echo "Sample is ready - all dependencies have been met"
|
||||
endif
|
||||
|
||||
main.o:src/main.cpp
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
pricingengine.o:src/pricingengine.cu
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
test.o:src/test.cpp
|
||||
$(EXEC) $(NVCC) $(INCLUDES) $(ALL_CCFLAGS) $(GENCODE_FLAGS) -o $@ -c $<
|
||||
|
||||
MC_SingleAsianOptionP: main.o pricingengine.o test.o
|
||||
$(EXEC) $(NVCC) $(ALL_LDFLAGS) $(GENCODE_FLAGS) -o $@ $+ $(LIBRARIES)
|
||||
$(EXEC) mkdir -p ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
$(EXEC) cp $@ ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)
|
||||
|
||||
run: build
|
||||
$(EXEC) ./MC_SingleAsianOptionP
|
||||
|
||||
testrun: build
|
||||
|
||||
clean:
|
||||
rm -f MC_SingleAsianOptionP main.o pricingengine.o test.o
|
||||
rm -rf ../../../bin/$(TARGET_ARCH)/$(TARGET_OS)/$(BUILD_TYPE)/MC_SingleAsianOptionP
|
||||
|
||||
clobber: clean
|
||||
@@ -0,0 +1,92 @@
|
||||
<?xml version="1.0" encoding="UTF-8"?>
|
||||
<!DOCTYPE entry SYSTEM "SamplesInfo.dtd">
|
||||
<entry>
|
||||
<name>MC_SingleAsianOptionP</name>
|
||||
<cuda_api_list>
|
||||
<toolkit>cudaFree</toolkit>
|
||||
<toolkit>cudaFuncGetAttributes</toolkit>
|
||||
<toolkit>cudaGetDeviceCount</toolkit>
|
||||
<toolkit>cudaMalloc</toolkit>
|
||||
<toolkit>cudaSetDevice</toolkit>
|
||||
<toolkit>cudaMemcpy</toolkit>
|
||||
<toolkit>cudaGetErrorString</toolkit>
|
||||
<toolkit>cudaGetDeviceProperties</toolkit>
|
||||
</cuda_api_list>
|
||||
<description><![CDATA[This sample uses Monte Carlo to simulate Single Asian Options using the NVIDIA CURAND library.]]></description>
|
||||
<devicecompilation>whole</devicecompilation>
|
||||
<includepaths>
|
||||
<path>./</path>
|
||||
<path>../</path>
|
||||
<path>../../../Common</path>
|
||||
</includepaths>
|
||||
<keyconcepts>
|
||||
<concept level="advanced">Random Number Generator</concept>
|
||||
<concept level="advanced">Computational Finance</concept>
|
||||
<concept level="advanced">CURAND Library</concept>
|
||||
</keyconcepts>
|
||||
<keywords>
|
||||
<keyword>CUDA</keyword>
|
||||
<keyword>CURAND</keyword>
|
||||
<keyword>GPGPU</keyword>
|
||||
<keyword>Monte Carlo</keyword>
|
||||
<keyword>rng</keyword>
|
||||
<keyword>prng</keyword>
|
||||
<keyword>qrng</keyword>
|
||||
<keyword>random number</keyword>
|
||||
</keywords>
|
||||
<libraries>
|
||||
<library>curand</library>
|
||||
</libraries>
|
||||
<librarypaths>
|
||||
</librarypaths>
|
||||
<nsight_eclipse>true</nsight_eclipse>
|
||||
<primary_file>src\main.cpp</primary_file>
|
||||
<required_dependencies>
|
||||
<dependency>CURAND</dependency>
|
||||
</required_dependencies>
|
||||
<scopes>
|
||||
<scope>1:CUDA Advanced Topics</scope>
|
||||
<scope>3:Computational Finance</scope>
|
||||
</scopes>
|
||||
<sm-arch>sm35</sm-arch>
|
||||
<sm-arch>sm37</sm-arch>
|
||||
<sm-arch>sm50</sm-arch>
|
||||
<sm-arch>sm52</sm-arch>
|
||||
<sm-arch>sm53</sm-arch>
|
||||
<sm-arch>sm60</sm-arch>
|
||||
<sm-arch>sm61</sm-arch>
|
||||
<sm-arch>sm70</sm-arch>
|
||||
<sm-arch>sm72</sm-arch>
|
||||
<sm-arch>sm75</sm-arch>
|
||||
<sm-arch>sm80</sm-arch>
|
||||
<sm-arch>sm86</sm-arch>
|
||||
<sm-arch>sm87</sm-arch>
|
||||
<supported_envs>
|
||||
<env>
|
||||
<arch>x86_64</arch>
|
||||
<platform>linux</platform>
|
||||
</env>
|
||||
<env>
|
||||
<platform>windows7</platform>
|
||||
</env>
|
||||
<env>
|
||||
<arch>x86_64</arch>
|
||||
<platform>macosx</platform>
|
||||
</env>
|
||||
<env>
|
||||
<arch>arm</arch>
|
||||
</env>
|
||||
<env>
|
||||
<arch>sbsa</arch>
|
||||
</env>
|
||||
<env>
|
||||
<arch>ppc64le</arch>
|
||||
<platform>linux</platform>
|
||||
</env>
|
||||
</supported_envs>
|
||||
<supported_sm_architectures>
|
||||
<include>all</include>
|
||||
</supported_sm_architectures>
|
||||
<title>Monte Carlo Single Asian Option</title>
|
||||
<type>exe</type>
|
||||
</entry>
|
||||
@@ -0,0 +1,74 @@
|
||||
# MC_SingleAsianOptionP - Monte Carlo Single Asian Option
|
||||
|
||||
## Description
|
||||
|
||||
This sample uses Monte Carlo to simulate Single Asian Options using the NVIDIA CURAND library.
|
||||
|
||||
## Key Concepts
|
||||
|
||||
Random Number Generator, Computational Finance, CURAND Library
|
||||
|
||||
## Supported SM Architectures
|
||||
|
||||
[SM 3.5 ](https://developer.nvidia.com/cuda-gpus) [SM 3.7 ](https://developer.nvidia.com/cuda-gpus) [SM 5.0 ](https://developer.nvidia.com/cuda-gpus) [SM 5.2 ](https://developer.nvidia.com/cuda-gpus) [SM 5.3 ](https://developer.nvidia.com/cuda-gpus) [SM 6.0 ](https://developer.nvidia.com/cuda-gpus) [SM 6.1 ](https://developer.nvidia.com/cuda-gpus) [SM 7.0 ](https://developer.nvidia.com/cuda-gpus) [SM 7.2 ](https://developer.nvidia.com/cuda-gpus) [SM 7.5 ](https://developer.nvidia.com/cuda-gpus) [SM 8.0 ](https://developer.nvidia.com/cuda-gpus) [SM 8.6 ](https://developer.nvidia.com/cuda-gpus) [SM 8.7 ](https://developer.nvidia.com/cuda-gpus)
|
||||
|
||||
## Supported OSes
|
||||
|
||||
Linux, Windows
|
||||
|
||||
## Supported CPU Architecture
|
||||
|
||||
x86_64, ppc64le, armv7l
|
||||
|
||||
## CUDA APIs involved
|
||||
|
||||
### [CUDA Runtime API](http://docs.nvidia.com/cuda/cuda-runtime-api/index.html)
|
||||
cudaFree, cudaFuncGetAttributes, cudaGetDeviceCount, cudaMalloc, cudaSetDevice, cudaMemcpy, cudaGetErrorString, cudaGetDeviceProperties
|
||||
|
||||
## Dependencies needed to build/run
|
||||
[CURAND](../../README.md#curand)
|
||||
|
||||
## Prerequisites
|
||||
|
||||
Download and install the [CUDA Toolkit 11.6](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
|
||||
Make sure the dependencies mentioned in [Dependencies]() section above are installed.
|
||||
|
||||
## Build and Run
|
||||
|
||||
### Windows
|
||||
The Windows samples are built using the Visual Studio IDE. Solution files (.sln) are provided for each supported version of Visual Studio, using the format:
|
||||
```
|
||||
*_vs<version>.sln - for Visual Studio <version>
|
||||
```
|
||||
Each individual sample has its own set of solution files in its directory:
|
||||
|
||||
To build/examine all the samples at once, the complete solution files should be used. To build/examine a single sample, the individual sample solution files should be used.
|
||||
> **Note:** Some samples require that the Microsoft DirectX SDK (June 2010 or newer) be installed and that the VC++ directory paths are properly set up (**Tools > Options...**). Check DirectX Dependencies section for details."
|
||||
|
||||
### Linux
|
||||
The Linux samples are built using makefiles. To use the makefiles, change the current directory to the sample directory you wish to build, and run make:
|
||||
```
|
||||
$ cd <sample_dir>
|
||||
$ make
|
||||
```
|
||||
The samples makefiles can take advantage of certain options:
|
||||
* **TARGET_ARCH=<arch>** - cross-compile targeting a specific architecture. Allowed architectures are x86_64, ppc64le, armv7l.
|
||||
By default, TARGET_ARCH is set to HOST_ARCH. On a x86_64 machine, not setting TARGET_ARCH is the equivalent of setting TARGET_ARCH=x86_64.<br/>
|
||||
`$ make TARGET_ARCH=x86_64` <br/> `$ make TARGET_ARCH=ppc64le` <br/> `$ make TARGET_ARCH=armv7l` <br/>
|
||||
See [here](http://docs.nvidia.com/cuda/cuda-samples/index.html#cross-samples) for more details.
|
||||
* **dbg=1** - build with debug symbols
|
||||
```
|
||||
$ make dbg=1
|
||||
```
|
||||
* **SMS="A B ..."** - override the SM architectures for which the sample will be built, where `"A B ..."` is a space-delimited list of SM architectures. For example, to generate SASS for SM 50 and SM 60, use `SMS="50 60"`.
|
||||
```
|
||||
$ make SMS="50 60"
|
||||
```
|
||||
|
||||
* **HOST_COMPILER=<host_compiler>** - override the default g++ host compiler. See the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#system-requirements) for a list of supported host compilers.
|
||||
```
|
||||
$ make HOST_COMPILER=g++
|
||||
```
|
||||
|
||||
## References (for more details)
|
||||
|
||||
@@ -0,0 +1,51 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#ifndef ASIANOPTION_H
|
||||
#define ASIANOPTION_H
|
||||
|
||||
template <typename Real>
|
||||
struct AsianOption {
|
||||
enum CallPut { Call, Put };
|
||||
|
||||
// Parameters
|
||||
Real spot;
|
||||
Real strike;
|
||||
Real r;
|
||||
Real sigma;
|
||||
Real tenor;
|
||||
Real dt;
|
||||
|
||||
// Value
|
||||
Real golden;
|
||||
Real value;
|
||||
|
||||
// Option type
|
||||
CallPut type;
|
||||
};
|
||||
|
||||
#endif
|
||||
@@ -0,0 +1,100 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#ifndef CUDASHAREDMEM_H
|
||||
#define CUDASHAREDMEM_H
|
||||
|
||||
//****************************************************************************
|
||||
// Because dynamically sized shared memory arrays are declared "extern",
|
||||
// we can't templatize them directly. To get around this, we declare a
|
||||
// simple wrapper struct that will declare the extern array with a different
|
||||
// name depending on the type. This avoids compiler errors about duplicate
|
||||
// definitions.
|
||||
//
|
||||
// To use dynamically allocated shared memory in a templatized __global__ or
|
||||
// __device__ function, just replace code like this:
|
||||
// template<class T>
|
||||
// __global__ void
|
||||
// foo( T* g_idata, T* g_odata)
|
||||
// {
|
||||
// // Shared mem size is determined by the host app at run time
|
||||
// extern __shared__ T sdata[];
|
||||
// ...
|
||||
// x = sdata[i];
|
||||
// sdata[i] = x;
|
||||
// ...
|
||||
// }
|
||||
//
|
||||
// With this:
|
||||
// template<class T>
|
||||
// __global__ void
|
||||
// foo( T* g_idata, T* g_odata)
|
||||
// {
|
||||
// // Shared mem size is determined by the host app at run time
|
||||
// SharedMemory<T> sdata;
|
||||
// ...
|
||||
// x = sdata[i];
|
||||
// sdata[i] = x;
|
||||
// ...
|
||||
// }
|
||||
//****************************************************************************
|
||||
|
||||
// This is the un-specialized struct. Note that we prevent instantiation of
|
||||
// this struct by making it abstract (i.e. with pure virtual methods).
|
||||
template <typename T>
|
||||
struct SharedMemory {
|
||||
// Ensure that we won't compile any un-specialized types
|
||||
virtual __device__ T &operator*() = 0;
|
||||
virtual __device__ T &operator[](int i) = 0;
|
||||
};
|
||||
|
||||
#define BUILD_SHAREDMEMORY_TYPE(t, n) \
|
||||
template <> \
|
||||
struct SharedMemory<t> { \
|
||||
__device__ t &operator*() { \
|
||||
extern __shared__ t n[]; \
|
||||
return *n; \
|
||||
} \
|
||||
__device__ t &operator[](int i) { \
|
||||
extern __shared__ t n[]; \
|
||||
return n[i]; \
|
||||
} \
|
||||
}
|
||||
|
||||
BUILD_SHAREDMEMORY_TYPE(int, s_int);
|
||||
BUILD_SHAREDMEMORY_TYPE(unsigned int, s_uint);
|
||||
BUILD_SHAREDMEMORY_TYPE(char, s_char);
|
||||
BUILD_SHAREDMEMORY_TYPE(unsigned char, s_uchar);
|
||||
BUILD_SHAREDMEMORY_TYPE(short, s_short);
|
||||
BUILD_SHAREDMEMORY_TYPE(unsigned short, s_ushort);
|
||||
BUILD_SHAREDMEMORY_TYPE(long, s_long);
|
||||
BUILD_SHAREDMEMORY_TYPE(unsigned long, s_ulong);
|
||||
BUILD_SHAREDMEMORY_TYPE(bool, s_bool);
|
||||
BUILD_SHAREDMEMORY_TYPE(float, s_float);
|
||||
BUILD_SHAREDMEMORY_TYPE(double, s_double);
|
||||
|
||||
#endif
|
||||
@@ -0,0 +1,47 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#ifndef PRICINGENGINE_H
|
||||
#define PRICINGENGINE_H
|
||||
|
||||
#include "asianoption.h"
|
||||
|
||||
template <typename Real>
|
||||
class PricingEngine {
|
||||
public:
|
||||
PricingEngine(unsigned int numSims, unsigned int device,
|
||||
unsigned int threadBlockSize, unsigned int seed);
|
||||
void operator()(AsianOption<Real> &option);
|
||||
|
||||
private:
|
||||
unsigned int m_seed;
|
||||
unsigned int m_numSims;
|
||||
unsigned int m_device;
|
||||
unsigned int m_threadBlockSize;
|
||||
};
|
||||
|
||||
#endif
|
||||
@@ -0,0 +1,57 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#ifndef TEST_H
|
||||
#define TEST_H
|
||||
|
||||
template <typename Real>
|
||||
struct Test {
|
||||
Test() : pass(false){};
|
||||
|
||||
int device;
|
||||
unsigned int numSims;
|
||||
unsigned int threadBlockSize;
|
||||
unsigned int seed;
|
||||
|
||||
bool pass;
|
||||
double elapsedTime;
|
||||
|
||||
bool operator()();
|
||||
};
|
||||
|
||||
// Defaults are arbitrary to give sensible runtime
|
||||
#define k_sims_min 100000
|
||||
#define k_sims_max 1000000
|
||||
#define k_sims_def 100000
|
||||
#define k_sims_qa 100000
|
||||
#define k_bsize_min 32
|
||||
#define k_bsize_def 128
|
||||
#define k_bsize_qa 128
|
||||
#define k_seed_def 1234
|
||||
#define k_seed_qa 1234
|
||||
|
||||
#endif
|
||||
@@ -0,0 +1,249 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
|
||||
///////////////////////////////////////////////////////////////////////////////
|
||||
// Monte Carlo: Single Asian Option
|
||||
// ================================
|
||||
//
|
||||
// This sample demonstrates pricing a single European-exercise average-price
|
||||
// Asian option (arithmetic discrete average).
|
||||
//
|
||||
// This file, main.cpp, contains the setup information to run the test, for
|
||||
// example parsing the command line and integrating this sample with the
|
||||
// samples framework. As such it is perhaps less interesting than the guts of
|
||||
// the sample. Readers wishing to skip the clutter are advised to skip straight
|
||||
// to Test.operator() in test.cpp.
|
||||
///////////////////////////////////////////////////////////////////////////////
|
||||
|
||||
|
||||
#include <iostream>
|
||||
#include <iomanip>
|
||||
#include <stdexcept>
|
||||
#include <cuda_runtime.h>
|
||||
#include <helper_timer.h>
|
||||
#include <helper_cuda.h>
|
||||
|
||||
#include "../inc/test.h"
|
||||
|
||||
// Forward declarations
|
||||
void showHelp(const int argc, const char **argv);
|
||||
template <typename Real>
|
||||
void runTest(int argc, const char **argv);
|
||||
|
||||
int main(int argc, char **argv) {
|
||||
using std::invalid_argument;
|
||||
using std::string;
|
||||
|
||||
// Open the log file
|
||||
printf("Monte Carlo Single Asian Option (with PRNG)\n");
|
||||
printf("===========================================\n\n");
|
||||
|
||||
// If help flag is set, display help and exit immediately
|
||||
if (checkCmdLineFlag(argc, (const char **)argv, "help")) {
|
||||
printf("Displaying help on console\n");
|
||||
showHelp(argc, (const char **)argv);
|
||||
exit(EXIT_SUCCESS);
|
||||
}
|
||||
|
||||
// Check the precision (checked against the device capability later)
|
||||
try {
|
||||
char *value;
|
||||
|
||||
if (getCmdLineArgumentString(argc, (const char **)argv, "precision",
|
||||
&value)) {
|
||||
// Check requested precision is valid
|
||||
string prec(value);
|
||||
|
||||
if (prec.compare("single") == 0 || prec.compare("\"single\"") == 0) {
|
||||
runTest<float>(argc, (const char **)argv);
|
||||
} else if (prec.compare("double") == 0 ||
|
||||
prec.compare("\"double\"") == 0) {
|
||||
runTest<double>(argc, (const char **)argv);
|
||||
} else {
|
||||
printf(
|
||||
"specified precision (%s) is invalid, must be \"single\" or "
|
||||
"\"double\".\n",
|
||||
value);
|
||||
throw invalid_argument("precision");
|
||||
}
|
||||
} else {
|
||||
runTest<float>(argc, (const char **)argv);
|
||||
}
|
||||
} catch (invalid_argument &e) {
|
||||
printf("invalid command line argument (%s)\n", e.what());
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
|
||||
// Finish
|
||||
exit(EXIT_SUCCESS);
|
||||
}
|
||||
|
||||
template <typename Real>
|
||||
void runTest(int argc, const char **argv) {
|
||||
using std::invalid_argument;
|
||||
using std::runtime_error;
|
||||
|
||||
try {
|
||||
Test<Real> test;
|
||||
int deviceCount = 0;
|
||||
cudaError_t cudaResult = cudaSuccess;
|
||||
|
||||
// by default specify GPU Device == 0
|
||||
test.device = 0;
|
||||
|
||||
// Get number of available devices
|
||||
cudaResult = cudaGetDeviceCount(&deviceCount);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
printf("could not get device count.\n");
|
||||
throw runtime_error("cudaGetDeviceCount");
|
||||
}
|
||||
|
||||
// (default parameters)
|
||||
test.numSims = k_sims_qa;
|
||||
test.threadBlockSize = k_bsize_qa;
|
||||
test.seed = k_seed_qa;
|
||||
|
||||
{
|
||||
char *value = 0;
|
||||
|
||||
if (getCmdLineArgumentString(argc, argv, "device", &value)) {
|
||||
test.device = (int)atoi(value);
|
||||
|
||||
if (test.device >= deviceCount) {
|
||||
printf(
|
||||
"invalid target device specified on command line (device %d does "
|
||||
"not exist).\n",
|
||||
test.device);
|
||||
throw invalid_argument("device");
|
||||
}
|
||||
} else {
|
||||
test.device = gpuGetMaxGflopsDeviceId();
|
||||
}
|
||||
|
||||
if (getCmdLineArgumentString(argc, argv, "sims", &value)) {
|
||||
test.numSims = (unsigned int)atoi(value);
|
||||
|
||||
if (test.numSims < k_sims_min || test.numSims > k_sims_max) {
|
||||
printf(
|
||||
"specified number of simulations (%d) is invalid, must be "
|
||||
"between %d and %d.\n",
|
||||
test.numSims, k_sims_min, k_sims_max);
|
||||
throw invalid_argument("sims");
|
||||
}
|
||||
} else {
|
||||
test.numSims = k_sims_def;
|
||||
}
|
||||
|
||||
if (getCmdLineArgumentString(argc, argv, "block-size", &value)) {
|
||||
// Determine max threads per block
|
||||
cudaDeviceProp deviceProperties;
|
||||
cudaResult = cudaGetDeviceProperties(&deviceProperties, test.device);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
printf("cound not get device properties for device %d.\n",
|
||||
test.device);
|
||||
throw runtime_error("cudaGetDeviceProperties");
|
||||
}
|
||||
|
||||
// Check requested size is valid
|
||||
test.threadBlockSize = (unsigned int)atoi(value);
|
||||
|
||||
if (test.threadBlockSize < k_bsize_min ||
|
||||
test.threadBlockSize > static_cast<unsigned int>(
|
||||
deviceProperties.maxThreadsPerBlock)) {
|
||||
printf(
|
||||
"specified block size (%d) is invalid, must be between %d and %d "
|
||||
"for device %d.\n",
|
||||
test.threadBlockSize, k_bsize_min,
|
||||
deviceProperties.maxThreadsPerBlock, test.device);
|
||||
throw invalid_argument("block-size");
|
||||
}
|
||||
|
||||
if (test.threadBlockSize & test.threadBlockSize - 1) {
|
||||
printf(
|
||||
"specified block size (%d) is invalid, must be a power of two "
|
||||
"(see reduction function).\n",
|
||||
test.threadBlockSize);
|
||||
throw invalid_argument("block-size");
|
||||
}
|
||||
} else {
|
||||
test.threadBlockSize = k_bsize_def;
|
||||
}
|
||||
|
||||
if (getCmdLineArgumentString(argc, argv, "seed", &value)) {
|
||||
// Check requested seed is valid
|
||||
test.seed = (unsigned int)atoi(value);
|
||||
|
||||
if (test.seed == 0) {
|
||||
printf("specified seed (%d) is invalid, must be non-zero.\n",
|
||||
test.seed);
|
||||
throw invalid_argument("seed");
|
||||
}
|
||||
} else {
|
||||
test.seed = k_seed_def;
|
||||
}
|
||||
}
|
||||
// Execute
|
||||
test();
|
||||
} catch (invalid_argument &e) {
|
||||
printf("invalid command line argument (%s)\n", e.what());
|
||||
exit(EXIT_FAILURE);
|
||||
} catch (runtime_error &e) {
|
||||
printf("runtime error (%s)\n", e.what());
|
||||
exit(EXIT_FAILURE);
|
||||
}
|
||||
}
|
||||
|
||||
void showHelp(int argc, const char **argv) {
|
||||
using std::cout;
|
||||
using std::endl;
|
||||
using std::left;
|
||||
using std::setw;
|
||||
|
||||
if (argc > 0) {
|
||||
cout << endl << argv[0] << endl;
|
||||
}
|
||||
|
||||
cout << endl << "Syntax:" << endl;
|
||||
cout << left;
|
||||
cout << " " << setw(20) << "--device=<device>"
|
||||
<< "Specify device to use for execution" << endl;
|
||||
cout << " " << setw(20) << "--sims=<N>"
|
||||
<< "Specify number of Monte Carlo simulations" << endl;
|
||||
cout << " " << setw(20) << "--block-size=<N>"
|
||||
<< "Specify number of threads per block" << endl;
|
||||
cout << " " << setw(20) << "--seed=<N>"
|
||||
<< "Specify the seed to use for the random number generator" << endl;
|
||||
cout << " " << setw(20) << "--precision=<P>"
|
||||
<< "Specify the precision (\"single\" or \"double\")" << endl;
|
||||
cout << endl;
|
||||
cout << " " << setw(20) << "--noprompt"
|
||||
<< "Skip prompt before exit" << endl;
|
||||
cout << endl;
|
||||
}
|
||||
@@ -0,0 +1,374 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#include "../inc/pricingengine.h"
|
||||
|
||||
#include <string>
|
||||
#include <vector>
|
||||
#include <numeric>
|
||||
#include <stdexcept>
|
||||
#include <typeinfo>
|
||||
#include <cuda_runtime.h>
|
||||
#include <cooperative_groups.h>
|
||||
|
||||
namespace cg = cooperative_groups;
|
||||
#include <curand_kernel.h>
|
||||
|
||||
#include "../inc/asianoption.h"
|
||||
#include "../inc/cudasharedmem.h"
|
||||
|
||||
using std::string;
|
||||
using std::vector;
|
||||
|
||||
// RNG init kernel
|
||||
__global__ void initRNG(curandState *const rngStates, const unsigned int seed) {
|
||||
// Determine thread ID
|
||||
unsigned int tid = blockIdx.x * blockDim.x + threadIdx.x;
|
||||
|
||||
// Initialise the RNG
|
||||
curand_init(seed, tid, 0, &rngStates[tid]);
|
||||
}
|
||||
|
||||
__device__ inline float getPathStep(float &drift, float &diffusion,
|
||||
curandState &state) {
|
||||
return expf(drift + diffusion * curand_normal(&state));
|
||||
}
|
||||
__device__ inline double getPathStep(double &drift, double &diffusion,
|
||||
curandState &state) {
|
||||
return exp(drift + diffusion * curand_normal_double(&state));
|
||||
}
|
||||
|
||||
// Path generation kernel
|
||||
template <typename Real>
|
||||
__global__ void generatePaths(Real *const paths, curandState *const rngStates,
|
||||
const AsianOption<Real> *const option,
|
||||
const unsigned int numSims,
|
||||
const unsigned int numTimesteps) {
|
||||
// Determine thread ID
|
||||
unsigned int tid = blockIdx.x * blockDim.x + threadIdx.x;
|
||||
unsigned int step = gridDim.x * blockDim.x;
|
||||
|
||||
// Compute parameters
|
||||
Real drift =
|
||||
(option->r - static_cast<Real>(0.5) * option->sigma * option->sigma) *
|
||||
option->dt;
|
||||
Real diffusion = option->sigma * sqrt(option->dt);
|
||||
|
||||
// Initialise the RNG
|
||||
curandState localState = rngStates[tid];
|
||||
|
||||
for (unsigned int i = tid; i < numSims; i += step) {
|
||||
// Shift the output pointer
|
||||
Real *output = paths + i;
|
||||
|
||||
// Simulate the path
|
||||
Real s = static_cast<Real>(1);
|
||||
|
||||
for (unsigned int t = 0; t < numTimesteps; t++, output += numSims) {
|
||||
s *= getPathStep(drift, diffusion, localState);
|
||||
*output = s;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
template <typename Real>
|
||||
__device__ Real reduce_sum(Real in, cg::thread_block cta) {
|
||||
SharedMemory<Real> sdata;
|
||||
|
||||
// Perform first level of reduction:
|
||||
// - Write to shared memory
|
||||
unsigned int ltid = threadIdx.x;
|
||||
|
||||
sdata[ltid] = in;
|
||||
cg::sync(cta);
|
||||
|
||||
// Do reduction in shared mem
|
||||
for (unsigned int s = blockDim.x / 2; s > 0; s >>= 1) {
|
||||
if (ltid < s) {
|
||||
sdata[ltid] += sdata[ltid + s];
|
||||
}
|
||||
|
||||
cg::sync(cta);
|
||||
}
|
||||
|
||||
return sdata[0];
|
||||
}
|
||||
|
||||
// Valuation kernel
|
||||
template <typename Real>
|
||||
__global__ void computeValue(Real *const values, const Real *const paths,
|
||||
const AsianOption<Real> *const option,
|
||||
const unsigned int numSims,
|
||||
const unsigned int numTimesteps) {
|
||||
// Handle to thread block group
|
||||
cg::thread_block cta = cg::this_thread_block();
|
||||
// Determine thread ID
|
||||
unsigned int bid = blockIdx.x;
|
||||
unsigned int tid = blockIdx.x * blockDim.x + threadIdx.x;
|
||||
unsigned int step = gridDim.x * blockDim.x;
|
||||
|
||||
Real sumPayoffs = static_cast<Real>(0);
|
||||
|
||||
for (unsigned int i = tid; i < numSims; i += step) {
|
||||
// Shift the input pointer
|
||||
const Real *path = paths + i;
|
||||
// Compute the arithmetic average
|
||||
Real avg = static_cast<Real>(0);
|
||||
|
||||
for (unsigned int t = 0; t < numTimesteps; t++, path += numSims) {
|
||||
avg += *path;
|
||||
}
|
||||
|
||||
avg = avg * option->spot / numTimesteps;
|
||||
// Compute the payoff
|
||||
Real payoff = avg - option->strike;
|
||||
|
||||
if (option->type == AsianOption<Real>::Put) {
|
||||
payoff = -payoff;
|
||||
}
|
||||
|
||||
payoff = max(static_cast<Real>(0), payoff);
|
||||
// Accumulate payoff locally
|
||||
sumPayoffs += payoff;
|
||||
}
|
||||
|
||||
// Reduce within the block
|
||||
sumPayoffs = reduce_sum<Real>(sumPayoffs, cta);
|
||||
|
||||
// Store the result
|
||||
if (threadIdx.x == 0) {
|
||||
values[bid] = sumPayoffs;
|
||||
}
|
||||
}
|
||||
|
||||
template <typename Real>
|
||||
PricingEngine<Real>::PricingEngine(unsigned int numSims, unsigned int device,
|
||||
unsigned int threadBlockSize,
|
||||
unsigned int seed)
|
||||
: m_numSims(numSims),
|
||||
m_device(device),
|
||||
m_threadBlockSize(threadBlockSize),
|
||||
m_seed(seed) {}
|
||||
|
||||
template <typename Real>
|
||||
void PricingEngine<Real>::operator()(AsianOption<Real> &option) {
|
||||
cudaError_t cudaResult = cudaSuccess;
|
||||
struct cudaDeviceProp deviceProperties;
|
||||
struct cudaFuncAttributes funcAttributes;
|
||||
|
||||
// Get device properties
|
||||
cudaResult = cudaGetDeviceProperties(&deviceProperties, m_device);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not get device properties: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Check precision is valid
|
||||
unsigned int deviceVersion =
|
||||
deviceProperties.major * 10 + deviceProperties.minor;
|
||||
|
||||
if (typeid(Real) == typeid(double) && deviceVersion < 13) {
|
||||
throw std::runtime_error("Device does not have double precision support");
|
||||
}
|
||||
|
||||
// Attach to GPU
|
||||
cudaResult = cudaSetDevice(m_device);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not set CUDA device: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Determine how to divide the work between cores
|
||||
dim3 block;
|
||||
dim3 grid;
|
||||
block.x = m_threadBlockSize;
|
||||
grid.x = (m_numSims + m_threadBlockSize - 1) / m_threadBlockSize;
|
||||
|
||||
// Aim to launch around ten or more times as many blocks as there
|
||||
// are multiprocessors on the target device.
|
||||
unsigned int blocksPerSM = 10;
|
||||
unsigned int numSMs = deviceProperties.multiProcessorCount;
|
||||
|
||||
while (grid.x > 2 * blocksPerSM * numSMs) {
|
||||
grid.x >>= 1;
|
||||
}
|
||||
|
||||
// Get initRNG function properties and check the maximum block size
|
||||
cudaResult = cudaFuncGetAttributes(&funcAttributes, initRNG);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not get function attributes: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
if (block.x > (unsigned int)funcAttributes.maxThreadsPerBlock) {
|
||||
throw std::runtime_error(
|
||||
"Block X dimension is too large for initRNG kernel");
|
||||
}
|
||||
|
||||
// Get generatePaths function properties and check the maximum block size
|
||||
cudaResult = cudaFuncGetAttributes(&funcAttributes, generatePaths<Real>);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not get function attributes: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
if (block.x > (unsigned int)funcAttributes.maxThreadsPerBlock) {
|
||||
throw std::runtime_error(
|
||||
"Block X dimension is too large for generatePaths kernel");
|
||||
}
|
||||
|
||||
// Get computeValue function properties and check the maximum block size
|
||||
cudaResult = cudaFuncGetAttributes(&funcAttributes, computeValue<Real>);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not get function attributes: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
if (block.x > (unsigned int)funcAttributes.maxThreadsPerBlock) {
|
||||
throw std::runtime_error(
|
||||
"Block X dimension is too large for computeValue kernel");
|
||||
}
|
||||
|
||||
// Setup problem on GPU
|
||||
AsianOption<Real> *d_option = 0;
|
||||
cudaResult = cudaMalloc((void **)&d_option, sizeof(AsianOption<Real>));
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not allocate memory on device for option data: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
cudaResult = cudaMemcpy(d_option, &option, sizeof(AsianOption<Real>),
|
||||
cudaMemcpyHostToDevice);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not copy data to device: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Allocate memory for paths
|
||||
Real *d_paths = 0;
|
||||
int numTimesteps = static_cast<int>(option.tenor / option.dt);
|
||||
cudaResult =
|
||||
cudaMalloc((void **)&d_paths, m_numSims * numTimesteps * sizeof(Real));
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not allocate memory on device for paths: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Allocate memory for RNG states
|
||||
curandState *d_rngStates = 0;
|
||||
cudaResult =
|
||||
cudaMalloc((void **)&d_rngStates, grid.x * block.x * sizeof(curandState));
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not allocate memory on device for RNG state: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Allocate memory for result
|
||||
Real *d_values = 0;
|
||||
cudaResult = cudaMalloc((void **)&d_values, grid.x * sizeof(Real));
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not allocate memory on device for partial results: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Initialise RNG
|
||||
initRNG<<<grid, block>>>(d_rngStates, m_seed);
|
||||
|
||||
// Generate paths
|
||||
generatePaths<Real><<<grid, block>>>(d_paths, d_rngStates, d_option,
|
||||
m_numSims, numTimesteps);
|
||||
|
||||
// Compute value
|
||||
computeValue<<<grid, block, block.x * sizeof(Real)>>>(
|
||||
d_values, d_paths, d_option, m_numSims, numTimesteps);
|
||||
|
||||
// Copy partial results back
|
||||
vector<Real> values(grid.x);
|
||||
cudaResult = cudaMemcpy(&values[0], d_values, grid.x * sizeof(Real),
|
||||
cudaMemcpyDeviceToHost);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
string msg("Could not copy partial results to host: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// Complete sum-reduction on host
|
||||
option.value =
|
||||
std::accumulate(values.begin(), values.end(), static_cast<Real>(0));
|
||||
|
||||
// Compute the mean
|
||||
option.value /= m_numSims;
|
||||
|
||||
// Discount to present value
|
||||
option.value *= exp(-option.r * option.tenor);
|
||||
|
||||
// Cleanup
|
||||
if (d_option) {
|
||||
cudaFree(d_option);
|
||||
d_option = 0;
|
||||
}
|
||||
|
||||
if (d_paths) {
|
||||
cudaFree(d_paths);
|
||||
d_paths = 0;
|
||||
}
|
||||
|
||||
if (d_rngStates) {
|
||||
cudaFree(d_rngStates);
|
||||
d_rngStates = 0;
|
||||
}
|
||||
|
||||
if (d_values) {
|
||||
cudaFree(d_values);
|
||||
d_values = 0;
|
||||
}
|
||||
}
|
||||
|
||||
// Explicit template instantiation
|
||||
template class PricingEngine<float>;
|
||||
template class PricingEngine<double>;
|
||||
@@ -0,0 +1,139 @@
|
||||
/* Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
|
||||
*
|
||||
* Redistribution and use in source and binary forms, with or without
|
||||
* modification, are permitted provided that the following conditions
|
||||
* are met:
|
||||
* * Redistributions of source code must retain the above copyright
|
||||
* notice, this list of conditions and the following disclaimer.
|
||||
* * Redistributions in binary form must reproduce the above copyright
|
||||
* notice, this list of conditions and the following disclaimer in the
|
||||
* documentation and/or other materials provided with the distribution.
|
||||
* * Neither the name of NVIDIA CORPORATION nor the names of its
|
||||
* contributors may be used to endorse or promote products derived
|
||||
* from this software without specific prior written permission.
|
||||
*
|
||||
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
|
||||
* EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
||||
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
||||
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||||
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
|
||||
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
|
||||
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
|
||||
* PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
|
||||
* OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
||||
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
||||
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
||||
*/
|
||||
|
||||
#include "../inc/test.h"
|
||||
|
||||
#include <sstream>
|
||||
#include <iomanip>
|
||||
#include <stdexcept>
|
||||
#include <memory>
|
||||
#include <iostream>
|
||||
#include <cassert>
|
||||
#include <typeinfo>
|
||||
#include <stdio.h>
|
||||
#include <cuda_runtime.h>
|
||||
#include <math.h>
|
||||
|
||||
#include <helper_timer.h>
|
||||
|
||||
#include "../inc/asianoption.h"
|
||||
#include "../inc/pricingengine.h"
|
||||
|
||||
template <typename Real>
|
||||
bool Test<Real>::operator()() {
|
||||
using std::endl;
|
||||
using std::setw;
|
||||
using std::stringstream;
|
||||
|
||||
StopWatchInterface *timer = NULL;
|
||||
sdkCreateTimer(&timer);
|
||||
|
||||
// Get device properties
|
||||
struct cudaDeviceProp deviceProperties;
|
||||
cudaError_t cudaResult = cudaGetDeviceProperties(&deviceProperties, device);
|
||||
|
||||
if (cudaResult != cudaSuccess) {
|
||||
std::string msg("Could not get device properties: ");
|
||||
msg += cudaGetErrorString(cudaResult);
|
||||
throw std::runtime_error(msg);
|
||||
}
|
||||
|
||||
// This test prices a single Asian call option with European
|
||||
// exercise, with the priced averaged arithmetically on discrete
|
||||
// trading days (weekdays).
|
||||
AsianOption<Real> option;
|
||||
option.spot = static_cast<Real>(40);
|
||||
option.strike = static_cast<Real>(35);
|
||||
option.r = static_cast<Real>(0.03);
|
||||
option.sigma = static_cast<Real>(0.20);
|
||||
option.tenor = static_cast<Real>(1.0 / 3.0);
|
||||
option.dt = static_cast<Real>(1.0 / 261);
|
||||
option.type = AsianOption<Real>::Call;
|
||||
option.value = static_cast<Real>(0.0);
|
||||
option.golden = static_cast<Real>(5.162534);
|
||||
|
||||
// Evaluate on GPU
|
||||
printf("Pricing option on GPU (%s)\n\n", deviceProperties.name);
|
||||
PricingEngine<Real> pricer(numSims, device, threadBlockSize, seed);
|
||||
sdkStartTimer(&timer);
|
||||
pricer(option);
|
||||
sdkStopTimer(&timer);
|
||||
elapsedTime = sdkGetAverageTimerValue(&timer) / 1000.0f;
|
||||
|
||||
// Tolerance to compare result with expected
|
||||
// This is just to check that nothing has gone very wrong with the
|
||||
// test, the actual accuracy of the result depends on the number of
|
||||
// Monte Carlo trials
|
||||
const Real tolerance = static_cast<Real>(0.1);
|
||||
|
||||
// Display results
|
||||
stringstream output;
|
||||
output << "Precision: "
|
||||
<< ((typeid(Real) == typeid(double)) ? "double" : "single") << endl;
|
||||
output << "Number of sims: " << numSims << endl;
|
||||
output << endl;
|
||||
output << " Spot | Strike | r | sigma | tenor | "
|
||||
"Call/Put | Value | Expected |"
|
||||
<< endl;
|
||||
output << "-----------|------------|------------|------------|------------|--"
|
||||
"----------|------------|------------|"
|
||||
<< endl;
|
||||
output << setw(10) << option.spot << " | ";
|
||||
output << setw(10) << option.strike << " | ";
|
||||
output << setw(10) << option.r << " | ";
|
||||
output << setw(10) << option.sigma << " | ";
|
||||
output << setw(10) << option.tenor << " | ";
|
||||
output << setw(10)
|
||||
<< (option.type == AsianOption<Real>::Call ? "Call" : "Put") << " | ";
|
||||
output << setw(10) << option.value << " | ";
|
||||
output << setw(10) << option.golden << " |";
|
||||
|
||||
printf("%s\n\n", output.str().c_str());
|
||||
|
||||
// Check result
|
||||
if (fabs(option.value - option.golden) > tolerance) {
|
||||
printf("computed result (%e) does not match expected result (%e).\n",
|
||||
option.value, option.golden);
|
||||
pass = false;
|
||||
} else {
|
||||
pass = true;
|
||||
}
|
||||
|
||||
// Print results
|
||||
printf(
|
||||
"MonteCarloSingleAsianOptionP, Performance = %.2f sims/s, Time = "
|
||||
"%.2f(ms), NumDevsUsed = %u, Blocksize = %u\n",
|
||||
numSims / elapsedTime, elapsedTime * 1000.0f, 1, threadBlockSize);
|
||||
|
||||
sdkDeleteTimer(&timer);
|
||||
|
||||
return pass;
|
||||
}
|
||||
|
||||
// Explicit template instantiation
|
||||
template struct Test<float>;
|
||||
template struct Test<double>;
|
||||
Reference in New Issue
Block a user