NvCloth 1.1.1 Release. (22392725)v1.1.1

author: Marijn Tamis <[email protected]> 2017-07-03 11:49:08 +0200
committer: Marijn Tamis <[email protected]> 2017-07-03 11:49:08 +0200
commit: cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381 (patch)
tree: 5cc014922d20561d87105d279b6f7eb3e628c6d9
parent: Fix windows line endings in github. (diff)
download: nvcloth-cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381.tar.xz
nvcloth-cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381.zip
58 files changed, 396 insertions, 544 deletions
diff --git a/NvCloth/CmakeGenerateProjects.bat b/NvCloth/CmakeGenerateProjects.bat
index a5891c9..fe50f54 100644
--- a/NvCloth/CmakeGenerateProjects.bat
+++ b/NvCloth/CmakeGenerateProjects.bat
@@ -30,38 +30,38 @@ REM Generate projects here
 rmdir /s /q compiler\vc11win32-cmake\
 mkdir compiler\vc11win32-cmake\
 pushd compiler\vc11win32-cmake\
-cmake ..\cmake\windows -G "Visual Studio 11 2012" -AWin32 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc11win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win32-cmake
+cmake ..\cmake\windows -G "Visual Studio 11 2012" -AWin32 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc11win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win32-cmake
 popd
 
 rmdir /s /q compiler\vc11win64-cmake\
 mkdir compiler\vc11win64-cmake\
 pushd compiler\vc11win64-cmake\
-cmake ..\cmake\windows -G "Visual Studio 11 2012" -Ax64 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc11win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win64-cmake
+cmake ..\cmake\windows -G "Visual Studio 11 2012" -Ax64 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc11win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc11win64-cmake
 popd
 
 
 rmdir /s /q compiler\vc12win32-cmake\
 mkdir compiler\vc12win32-cmake\
 pushd compiler\vc12win32-cmake\
-cmake ..\cmake\windows -G "Visual Studio 12 2013" -AWin32 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc12win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win32-cmake
+cmake ..\cmake\windows -G "Visual Studio 12 2013" -AWin32 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc12win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win32-cmake
 popd
 
 rmdir /s /q compiler\vc12win64-cmake\
 mkdir compiler\vc12win64-cmake\
 pushd compiler\vc12win64-cmake\
-cmake ..\cmake\windows -G "Visual Studio 12 2013" -Ax64 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc12win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win64-cmake
+cmake ..\cmake\windows -G "Visual Studio 12 2013" -Ax64 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc12win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc12win64-cmake
 popd
 
 rmdir /s /q compiler\vc14win32-cmake\
 mkdir compiler\vc14win32-cmake\
 pushd compiler\vc14win32-cmake\
-cmake ..\cmake\windows -G "Visual Studio 14 2015" -AWin32 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc14win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win32-cmake
+cmake ..\cmake\windows -G "Visual Studio 14 2015" -AWin32 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win32-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc14win32-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win32-cmake
 popd
 
 rmdir /s /q compiler\vc14win64-cmake\
 mkdir compiler\vc14win64-cmake\
 pushd compiler\vc14win64-cmake\
-cmake ..\cmake\windows -G "Visual Studio 14 2015" -Ax64 -DTARGET_BUILD_PLATFORM=Windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc14win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win64-cmake
+cmake ..\cmake\windows -G "Visual Studio 14 2015" -Ax64 -DTARGET_BUILD_PLATFORM=windows -DPX_GENERATE_GPU_PROJECTS=1 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DSTATIC_WINCRT=1 -DPX_OUTPUT_DLL_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win64-cmake -DPX_OUTPUT_LIB_DIR=%PX_OUTPUT_ROOT%\Lib\vc14win64-cmake -DPX_OUTPUT_EXE_DIR=%PX_OUTPUT_ROOT%\Bin\vc14win64-cmake
 popd
 
 
diff --git a/NvCloth/GenerateProjectsLinux.sh b/NvCloth/GenerateProjectsLinux.sh
index e34e51f..56d1788 100644
--- a/NvCloth/GenerateProjectsLinux.sh
+++ b/NvCloth/GenerateProjectsLinux.sh
@@ -19,23 +19,23 @@ export PX_OUTPUT_ROOT="$PWD"
 rm -r -f compiler/linux64-debug-cmake/
 mkdir compiler/linux64-debug-cmake/
 cd compiler/linux64-debug-cmake/
-cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=Linux -DCMAKE_BUILD_TYPE=debug -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
+cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=linux -DCMAKE_BUILD_TYPE=debug -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
 cd ../../
 
 rm -r -f compiler/linux64-checked-cmake/
 mkdir compiler/linux64-checked-cmake/
 cd compiler/linux64-checked-cmake/
-cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=Linux -DCMAKE_BUILD_TYPE=checked -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
+cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=linux -DCMAKE_BUILD_TYPE=checked -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
 cd ../../
 
 rm -r -f compiler/linux64-profile-cmake/
 mkdir compiler/linux64-profile-cmake/
 cd compiler/linux64-profile-cmake/
-cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=Linux -DCMAKE_BUILD_TYPE=profile -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
+cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=linux -DCMAKE_BUILD_TYPE=profile -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
 cd ../../
 
 rm -r -f compiler/linux64-release-cmake/
 mkdir compiler/linux64-release-cmake/
 cd compiler/linux64-release-cmake/
-cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=Linux -DCMAKE_BUILD_TYPE=release -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
+cmake ../cmake/Linux -G "Unix Makefiles" -DTARGET_BUILD_PLATFORM=linux -DCMAKE_BUILD_TYPE=release -DPX_GENERATE_GPU_PROJECTS=0 -DPX_OUTPUT_DLL_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/lib/linux64-cmake -DPX_OUTPUT_EXE_DIR=$PX_OUTPUT_ROOT/bin/linux64-cmake
 cd ../../
 \ No newline at end of file
diff --git a/NvCloth/GenerateProjectsOsx.sh b/NvCloth/GenerateProjectsOsx.sh
index 5f0e3c7..5c57f42 100644
--- a/NvCloth/GenerateProjectsOsx.sh
+++ b/NvCloth/GenerateProjectsOsx.sh
@@ -19,13 +19,13 @@ export PX_OUTPUT_ROOT="$PWD"
 rm -r -f compiler/osx32-cmake/
 mkdir compiler/osx32-cmake/
 cd compiler/osx32-cmake/
-cmake ../cmake/Mac -G Xcode -DTARGET_BUILD_PLATFORM=Mac -DPX_32BIT=1 -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/Lib/osx32-cmake 
+cmake ../cmake/Mac -G Xcode -DTARGET_BUILD_PLATFORM=mac -DPX_32BIT=1 -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/Lib/osx32-cmake 
 cd ../../
 
 rm -r -f compiler/osx64-cmake/
 mkdir compiler/osx64-cmake/
 cd compiler/osx64-cmake/
-cmake ../cmake/Mac -G Xcode -DTARGET_BUILD_PLATFORM=Mac -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/Lib/osx64-cmake 
+cmake ../cmake/Mac -G Xcode -DTARGET_BUILD_PLATFORM=mac -DPX_OUTPUT_LIB_DIR=$PX_OUTPUT_ROOT/Lib/osx64-cmake 
 cd ../../
 
 
diff --git a/NvCloth/ReleaseNotes.txt b/NvCloth/ReleaseNotes.txt
index 5751a25..7dfa9c0 100644
--- a/NvCloth/ReleaseNotes.txt
+++ b/NvCloth/ReleaseNotes.txt
@@ -1,9 +1,9 @@
 NvCloth Release notes.
 
-[1.1.0]
+[1.1.1]
 Supported platforms:
-* Windows (CPU, CUDA, DX11, RenderCloth)
-* Mac (beta, tested on Sierra, RenderCloth)
+* Windows (CPU, CUDA, DX11)
+* Mac (beta, tested on Sierra)
 * Linux (beta, tested on Ubuntu 16.04.1 LTS x64)
 * PlayStation 4 (beta) with PS4 SDK 4.5
 * Xbox one (CPU, DX11, beta)
@@ -17,6 +17,15 @@ Supported compilers (via cmake 3.7)
 * Xbox one: Visual studio 14 (2015)
 
 Fixed:
+Air drag/lift did react to flow speed only linearly, model was changed to be more physically correct.
+Fixed issue where scaling units would result in different air drag/lift behavior (by adding fluid density parameter).
+Fixed "unused typedef `__t100`" warning in xcode.
+
+Known issues:
+DirectX 11 solver has issues with ccd sphere/capsule collision.
+
+[1.1.0]
+Fixed:
 DirectX 11 solver now supports triangle collision shapes and air drag/lift.
 Fixed bug where changing friction on cloth already in simulation would cause assert or crash.
 Optimization flag now working on PS4 release build.
diff --git a/NvCloth/compiler/cmake/Linux/CMakeLists.txt b/NvCloth/compiler/cmake/Linux/CMakeLists.txt
index 4d8d85a..9425dde 100644
--- a/NvCloth/compiler/cmake/Linux/CMakeLists.txt
+++ b/NvCloth/compiler/cmake/Linux/CMakeLists.txt
@@ -4,10 +4,10 @@ include(../common/CMakeLists.txt)
 
 
 IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Linux
-	SET(TARGET_BUILD_PLATFORM "Linux")
+	SET(TARGET_BUILD_PLATFORM "linux")
 ENDIF()
 
-SET(PLATFORM_LIST Linux)
+SET(PLATFORM_LIST linux)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
@@ -47,7 +47,7 @@ FIND_PACKAGE(PxShared REQUIRED)
 #  such as - we don't want to be installing any built artifacts on this server, etc. So for now we hack it.
 
 # Add PxShared as a dependency so that we can use project references
-ADD_SUBDIRECTORY(${PXSHARED_ROOT_DIR}/src/compiler/cmake/Linux "${CMAKE_CURRENT_BINARY_DIR}/pxshared_bin")
+ADD_SUBDIRECTORY(${PXSHARED_ROOT_DIR}/src/compiler/cmake/linux "${CMAKE_CURRENT_BINARY_DIR}/pxshared_bin")
 
 # Include all of the projects
 INCLUDE(NvCloth.cmake)
diff --git a/NvCloth/compiler/cmake/Mac/CMakeLists.txt b/NvCloth/compiler/cmake/Mac/CMakeLists.txt
index 05a42ff..4e1c124 100644
--- a/NvCloth/compiler/cmake/Mac/CMakeLists.txt
+++ b/NvCloth/compiler/cmake/Mac/CMakeLists.txt
@@ -4,10 +4,10 @@ include(../common/CMakeLists.txt)
 
 
 IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Mac
-	SET(TARGET_BUILD_PLATFORM "Mac")
+	SET(TARGET_BUILD_PLATFORM "mac")
 ENDIF()
 
-SET(PLATFORM_LIST Mac)
+SET(PLATFORM_LIST mac)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
diff --git a/NvCloth/compiler/cmake/windows/CMakeLists.txt b/NvCloth/compiler/cmake/windows/CMakeLists.txt
index 79b0236..81d8289 100644
--- a/NvCloth/compiler/cmake/windows/CMakeLists.txt
+++ b/NvCloth/compiler/cmake/windows/CMakeLists.txt
@@ -4,10 +4,10 @@ include(../common/CMakeLists.txt)
 
 
 IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Windows
-	SET(TARGET_BUILD_PLATFORM "Windows")
+	SET(TARGET_BUILD_PLATFORM "windows")
 ENDIF()
 
-SET(PLATFORM_LIST Windows)
+SET(PLATFORM_LIST windows)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
diff --git a/NvCloth/extensions/src/ClothFabricCooker.cpp b/NvCloth/extensions/src/ClothFabricCooker.cpp
index f0e4dea..2d800e9 100644
--- a/NvCloth/extensions/src/ClothFabricCooker.cpp
+++ b/NvCloth/extensions/src/ClothFabricCooker.cpp
@@ -97,7 +97,9 @@ nv::cloth::Range<const T> CreateRange(typename nv::cloth::Vector<T>::Type const&
 template<typename T, typename U>
 nv::cloth::Range<const T> CreateRangeF(typename nv::cloth::Vector<U>::Type const& vector, int offset = 0)
 {
+#ifndef _LIBCPP_HAS_NO_STATIC_ASSERT
 	static_assert(sizeof(T) == sizeof(U), "Type T and U need to be of the same size");
+#endif
 	const T* begin = reinterpret_cast<const T*>(vector.begin()+offset);
 	const T* end = reinterpret_cast<const T*>(vector.end());
 
diff --git a/NvCloth/include/NvCloth/Cloth.h b/NvCloth/include/NvCloth/Cloth.h
index 3d783e9..ac957f7 100644
--- a/NvCloth/include/NvCloth/Cloth.h
+++ b/NvCloth/include/NvCloth/Cloth.h
@@ -376,6 +376,10 @@ class Cloth : public UserAllocated
 	virtual void setLiftCoefficient(float) = 0;
 	///Returns value set with setLiftCoefficient().
 	virtual float getLiftCoefficient() const = 0;
+	/** /brief Sets the fluid density used for air drag/lift calculations. */
+	virtual void setFluidDensity(float) = 0;
+	///Returns value set with setFluidDensity().
+	virtual float getFluidDensity() const = 0;
 
 	/* self collision */
 
diff --git a/NvCloth/samples/CmakeGenerateProjects.bat b/NvCloth/samples/CmakeGenerateProjects.bat
index d63cde8..8c6d88c 100644
--- a/NvCloth/samples/CmakeGenerateProjects.bat
+++ b/NvCloth/samples/CmakeGenerateProjects.bat
@@ -14,6 +14,9 @@ SET PATH=%PATH%;"%CMAKE_PATH_F%"
 
 REM Make sure the various variables that we need are set
 
+call "../scripts/locate_cuda.bat" CUDA_PATH_
+echo CUDA_PATH_ = %CUDA_PATH_%
+
 IF EXIST %~dp0..\Externals\CMakeModules (
 	set GW_DEPS_ROOT=%~dp0..\
 )
@@ -30,13 +33,13 @@ REM Generate projects here
 rmdir /s /q compiler\vc14win32-cmake\
 mkdir compiler\vc14win32-cmake\
 pushd compiler\vc14win32-cmake\
-cmake ..\.. -G "Visual Studio 14 2015" -AWin32 -DTARGET_BUILD_PLATFORM=Windows -DSTATIC_WINCRT=0 -DBL_DLL_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win32-cmake -DBL_LIB_OUTPUT_DIR=%OUTPUT_ROOT%\lib\vc14win32-cmake -DBL_EXE_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win32-cmake
+cmake ..\.. -G "Visual Studio 14 2015" -AWin32 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DTARGET_BUILD_PLATFORM=windows -DSTATIC_WINCRT=0 -DBL_DLL_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win32-cmake -DBL_LIB_OUTPUT_DIR=%OUTPUT_ROOT%\lib\vc14win32-cmake -DBL_EXE_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win32-cmake
 popd
 
 rmdir /s /q compiler\vc14win64-cmake\
 mkdir compiler\vc14win64-cmake\
 pushd compiler\vc14win64-cmake\
-cmake ..\.. -G "Visual Studio 14 2015" -Ax64 -DTARGET_BUILD_PLATFORM=Windows -DSTATIC_WINCRT=0 -DBL_DLL_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win64-cmake -DBL_LIB_OUTPUT_DIR=%OUTPUT_ROOT%\lib\vc14win64-cmake -DBL_EXE_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win64-cmake
+cmake ..\.. -G "Visual Studio 14 2015" -Ax64 -DCUDA_TOOLKIT_ROOT_DIR="%CUDA_PATH_%" -DTARGET_BUILD_PLATFORM=windows -DSTATIC_WINCRT=0 -DBL_DLL_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win64-cmake -DBL_LIB_OUTPUT_DIR=%OUTPUT_ROOT%\lib\vc14win64-cmake -DBL_EXE_OUTPUT_DIR=%OUTPUT_ROOT%\bin\vc14win64-cmake
 popd
 
 
diff --git a/NvCloth/samples/SampleBase/Main.cpp b/NvCloth/samples/SampleBase/Main.cpp
index 433ed88..f4107c3 100644
--- a/NvCloth/samples/SampleBase/Main.cpp
+++ b/NvCloth/samples/SampleBase/Main.cpp
@@ -12,7 +12,9 @@
 #include <sstream>
 
 #include <Windows.h>
-
+#include <iostream>
+#include <io.h>
+#include <fcntl.h>
 
 using namespace std;
 
@@ -22,6 +24,8 @@ int WINAPI wWinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, LPWSTR lpCmdLi
 #if defined(DEBUG) | defined(_DEBUG)
 	_CrtSetDbgFlag(_CRTDBG_ALLOC_MEM_DF | _CRTDBG_LEAK_CHECK_DF);
 	AllocConsole();
+	FILE* fp;
+	freopen_s(&fp, "CONOUT$", "w", stdout);
 #endif
 
 	SampleConfig config;
diff --git a/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.cpp b/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.cpp
new file mode 100644
index 0000000..f2a5b82
--- /dev/null
+++ b/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.cpp
@@ -0,0 +1,98 @@
+/*
+* Copyright (c) 2008-2017, NVIDIA CORPORATION.  All rights reserved.
+*
+* NVIDIA CORPORATION and its licensors retain all intellectual property
+* and proprietary rights in and to this software, related documentation
+* and any modifications thereto.  Any use, reproduction, disclosure or
+* distribution of this software and related documentation without an express
+* license agreement from NVIDIA CORPORATION is strictly prohibited.
+*/
+
+#include "ScaledScene.h"
+#include "Scene/SceneController.h"
+#include <NvClothExt/ClothFabricCooker.h>
+#include "ClothMeshGenerator.h"
+#include <NvCloth/Fabric.h>
+#include <NvCloth/Solver.h>
+#include <NvCloth/Cloth.h>
+#include <NvCloth/Factory.h>
+#include "Renderer.h"
+#include "renderer/RenderUtils.h"
+ 
+DECLARE_SCENE_NAME(ScaledScene,"Scaled Scene")
+
+void ScaledScene::onInitialize()
+{
+	///////////////////////////////////////////////////////////////////////
+	ClothMeshData clothMesh;
+
+	physx::PxMat44 transform = PxTransform(PxVec3(-2.f, 13.f, 0.f)*0.0, PxQuat(PxPi / 6.f, PxVec3(1.f, 0.f, 0.f)));
+	clothMesh.GeneratePlaneCloth(600.f, 700.f, 49, 59, false, transform);
+	clothMesh.AttachClothPlaneByAngles(49, 59);
+
+	mClothActor = new ClothActor;
+	nv::cloth::ClothMeshDesc meshDesc = clothMesh.GetClothMeshDesc();
+	{
+		mClothActor->mClothRenderMesh = new ClothRenderMesh(meshDesc);
+		mClothActor->mClothRenderable = getSceneController()->getRenderer().createRenderable(*(static_cast<IRenderMesh*>(mClothActor->mClothRenderMesh)), *getSceneController()->getDefaultMaterial());
+		mClothActor->mClothRenderable->setColor(getRandomPastelColor());
+		mClothActor->mClothRenderable->setScale(physx::PxVec3(0.01, 0.01, 0.01));
+		mClothActor->mClothRenderable->setTransform(PxTransform(PxVec3(-2.f, 13.f, 0.f),physx::PxQuat(physx::PxIdentity)));
+	}
+
+	nv::cloth::Vector<int32_t>::Type phaseTypeInfo;
+	mFabric = NvClothCookFabricFromMesh(getSceneController()->getFactory(), meshDesc, physx::PxVec3(0.0f, -9.8f, 0.0f), &phaseTypeInfo, false);
+	trackFabric(mFabric);
+
+	// Initialize start positions and masses for the actual cloth instance
+	// (note: the particle/vertex positions do not have to match the mesh description here. Set the positions to the initial shape of this cloth instance)
+	std::vector<physx::PxVec4> particlesCopy;
+	particlesCopy.resize(clothMesh.mVertices.size());
+
+	physx::PxVec3 center = transform.transform(physx::PxVec3(0.0f, 0.0f, 0.0f));
+	for (int i = 0; i < (int)clothMesh.mVertices.size(); i++)
+	{
+		// To put attachment point closer to each other
+		if(clothMesh.mInvMasses[i] < 1e-6)
+			clothMesh.mVertices[i] = (clothMesh.mVertices[i] - center) * 0.85f + center;
+
+		particlesCopy[i] = physx::PxVec4(clothMesh.mVertices[i], clothMesh.mInvMasses[i]); // w component is 1/mass, or 0.0f for anchored/fixed particles
+	}
+
+	// Create the cloth from the initial positions/masses and the fabric
+	mClothActor->mCloth = getSceneController()->getFactory()->createCloth(nv::cloth::Range<physx::PxVec4>(&particlesCopy[0], &particlesCopy[0] + particlesCopy.size()), *mFabric);
+	particlesCopy.clear(); particlesCopy.shrink_to_fit();
+
+	mClothActor->mCloth->setGravity(physx::PxVec3(0.0f, -980.0f, 0.0f));
+
+	// Setup phase configs
+	std::vector<nv::cloth::PhaseConfig> phases(mFabric->getNumPhases());
+	for (int i = 0; i < (int)phases.size(); i++)
+	{
+		phases[i].mPhaseIndex = i;
+		phases[i].mStiffness = 1.0f;
+		phases[i].mStiffnessMultiplier = 1.0f;
+		phases[i].mCompressionLimit = 1.0f;
+		phases[i].mStretchLimit = 1.0f;
+	}
+	mClothActor->mCloth->setPhaseConfig(nv::cloth::Range<nv::cloth::PhaseConfig>(&phases.front(), &phases.back()));
+	mClothActor->mCloth->setDragCoefficient(0.1f);
+	mClothActor->mCloth->setLiftCoefficient(0.1f);
+	//mClothActor->mCloth->setWindVelocity(physx::PxVec3(50, 0.0, 50.0));
+	mClothActor->mCloth->setFluidDensity(1.0f / powf(100, 3));
+
+	mSolver = getSceneController()->getFactory()->createSolver();
+	trackSolver(mSolver);
+	trackClothActor(mClothActor);
+
+	// Add the cloth to the solver for simulation
+	addClothToSolver(mClothActor, mSolver);
+	
+	{
+		IRenderMesh* mesh = getSceneController()->getRenderer().getPrimitiveRenderMesh(PrimitiveRenderMeshType::Plane);
+		Renderable* plane = getSceneController()->getRenderer().createRenderable(*mesh, *getSceneController()->getDefaultPlaneMaterial());
+		plane->setTransform(PxTransform(PxVec3(0.f, 0.f, 0.f), PxQuat(PxPiDivTwo, PxVec3(0.f, 0.f, 1.f))));
+		plane->setScale(PxVec3(1000.f));
+		trackRenderable(plane);
+	}
+}
diff --git a/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.h b/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.h
new file mode 100644
index 0000000..eb47c36
--- /dev/null
+++ b/NvCloth/samples/SampleBase/scene/scenes/ScaledScene.h
@@ -0,0 +1,32 @@
+/*
+* Copyright (c) 2008-2017, NVIDIA CORPORATION.  All rights reserved.
+*
+* NVIDIA CORPORATION and its licensors retain all intellectual property
+* and proprietary rights in and to this software, related documentation
+* and any modifications thereto.  Any use, reproduction, disclosure or
+* distribution of this software and related documentation without an express
+* license agreement from NVIDIA CORPORATION is strictly prohibited.
+*/
+
+#ifndef SCALED_SCENE_H
+#define SCALED_SCENE_H
+
+#include "scene/Scene.h"
+
+class ScaledScene : public Scene
+{
+public:
+
+	ScaledScene(SceneController* sceneController):Scene(sceneController) {}
+
+	virtual void onInitialize() override;
+
+private:
+	nv::cloth::Fabric* mFabric;
+	nv::cloth::Solver* mSolver;
+	ClothActor* mClothActor;
+
+};
+
+
+#endif
+\ No newline at end of file
diff --git a/NvCloth/samples/SampleBase/scene/scenes/WindScene.cpp b/NvCloth/samples/SampleBase/scene/scenes/WindScene.cpp
index 232316d..d84fb87 100644
--- a/NvCloth/samples/SampleBase/scene/scenes/WindScene.cpp
+++ b/NvCloth/samples/SampleBase/scene/scenes/WindScene.cpp
@@ -35,7 +35,7 @@ void WindScene::Animate(double dt)
 
 		for(int i = 0; i < 3; i++)
 		{
-			physx::PxVec3 wind = physx::PxVec3(dvx, vy, dvz);
+			physx::PxVec3 wind = physx::PxVec3(dvx, vy, dvz)/5.0f;
 			mClothActor[i]->mCloth->setWindVelocity(wind);
 		}
 	}
diff --git a/NvCloth/samples/compiler/cmake/SampleBase.cmake b/NvCloth/samples/compiler/cmake/SampleBase.cmake
index e3aa2cc..cee4779 100644
--- a/NvCloth/samples/compiler/cmake/SampleBase.cmake
+++ b/NvCloth/samples/compiler/cmake/SampleBase.cmake
@@ -110,6 +110,8 @@ SET(SCENES_FILES
 	${SB_SCENE_SOURCE_DIR}/scenes/CCDScene.h
 	${SB_SCENE_SOURCE_DIR}/scenes/MultiSolverScene.cpp
 	${SB_SCENE_SOURCE_DIR}/scenes/MultiSolverScene.h
+	${SB_SCENE_SOURCE_DIR}/scenes/ScaledScene.cpp
+	${SB_SCENE_SOURCE_DIR}/scenes/ScaledScene.h
 )
 
 SET(UI_FILES
diff --git a/NvCloth/src/ClothBase.h b/NvCloth/src/ClothBase.h
index 8d75a72..ec1ee40 100644
--- a/NvCloth/src/ClothBase.h
+++ b/NvCloth/src/ClothBase.h
@@ -74,6 +74,7 @@ void initialize(Cloth& cloth, const physx::PxVec4* pIt, const physx::PxVec4* pEn
 	cloth.mWind = physx::PxVec3(0.0f);
 	cloth.mDragLogCoefficient = 0.0f;
 	cloth.mLiftLogCoefficient = 0.0f;
+	cloth.mFluidDensity = 1.0f;
 	cloth.mEnableContinuousCollision = false;
 	cloth.mCollisionMassScale = 0.0f;
 	cloth.mFriction = 0.0f;
diff --git a/NvCloth/src/ClothImpl.h b/NvCloth/src/ClothImpl.h
index 0c2b362..1e8d9a1 100644
--- a/NvCloth/src/ClothImpl.h
+++ b/NvCloth/src/ClothImpl.h
@@ -157,6 +157,8 @@ class ClothImpl : public Cloth
 	virtual float getDragCoefficient() const;
 	virtual void setLiftCoefficient(float);
 	virtual float getLiftCoefficient() const;
+	virtual void setFluidDensity(float);
+	virtual float getFluidDensity() const;
 
 	virtual void setSelfCollisionDistance(float);
 	virtual float getSelfCollisionDistance() const;
@@ -216,6 +218,7 @@ public: //Fields shared between all cloth classes
 	physx::PxVec3 mWind;
 	float mDragLogCoefficient;
 	float mLiftLogCoefficient;
+	float mFluidDensity;
 
 	// sleeping
 	uint32_t mSleepTestInterval; // how often to test for movement
@@ -313,7 +316,8 @@ inline physx::PxVec3 ClothImpl<T>::getGravity() const
 
 inline float safeLog2(float x)
 {
-	return x ? physx::shdfnd::log2(x) : -FLT_MAX_EXP;
+	NV_CLOTH_ASSERT(("safeLog2",x >= 0.0f));
+	return x > 0 ? physx::shdfnd::log2(x) : -FLT_MAX_EXP;
 }
 
 inline physx::PxVec3 safeLog2(const physx::PxVec3& v)
@@ -1214,11 +1218,30 @@ inline float ClothImpl<T>::getLiftCoefficient() const
 }
 
 template <typename T>
+inline void ClothImpl<T>::setFluidDensity(float fluidDensity)
+{
+	NV_CLOTH_ASSERT(fluidDensity < 0.f);
+	if (fluidDensity == mFluidDensity)
+		return;
+
+	mFluidDensity = fluidDensity;
+	getChildCloth()->notifyChanged();
+	wakeUp();
+}
+
+template <typename T>
+inline float ClothImpl<T>::getFluidDensity() const
+{
+	return mFluidDensity;
+}
+
+template <typename T>
 inline uint32_t ClothImpl<T>::getNumSelfCollisionIndices() const
 {
 	return uint32_t(getChildCloth()->mSelfCollisionIndices.size());
 }
 
+
 // Fixed 4505:local function has been removed
 template <typename T>
 inline void ClothImpl<T>::setRestPositions(Range<const physx::PxVec4> restPositions)
@@ -1255,6 +1278,7 @@ inline float ClothImpl<T>::getSelfCollisionDistance() const
 template <typename T>
 inline void ClothImpl<T>::setSelfCollisionStiffness(float stiffness)
 {
+	NV_CLOTH_ASSERT(stiffness <= 1.0f);
 	float value = safeLog2(1 - stiffness);
 	if (value == getChildCloth()->mSelfCollisionLogStiffness)
 		return;
diff --git a/NvCloth/src/IterationState.h b/NvCloth/src/IterationState.h
index f199663..224e87e 100644
--- a/NvCloth/src/IterationState.h
+++ b/NvCloth/src/IterationState.h
@@ -137,7 +137,7 @@ struct IterationState
 
 	Simd4f mCurBias;  // in local space
 	Simd4f mPrevBias; // in local space
-	Simd4f mWind;     // delta position per iteration
+	Simd4f mWind;     // delta position per iteration (wind velocity * mIterDt)
 
 	Simd4f mPrevMatrix[3];
 	Simd4f mCurMatrix[3];
@@ -290,7 +290,7 @@ cloth::IterationState<Simd4f> cloth::IterationStateFactory::create(MyCloth const
 	result.mCurBias = transform(result.mRotationMatrix, curLinearInertia + bias) & maskXYZ;
 	result.mPrevBias = transform(result.mRotationMatrix, linearInertia - curLinearInertia) & maskXYZ;
 
-	Simd4f wind = load(array(cloth.mWind)) * iterDt;
+	Simd4f wind = load(array(cloth.mWind)) * iterDt; // multiply with delta time here already so we don't have to do it inside the solver
 	result.mWind = transform(result.mRotationMatrix, translation - wind) & maskXYZ;
 
 	result.mIsTurning = mPrevAngularVelocity.magnitudeSquared() + cloth.mAngularVelocity.magnitudeSquared() > 0.0f;
diff --git a/NvCloth/src/SwClothData.cpp b/NvCloth/src/SwClothData.cpp
index eddd821..f102bde 100644
--- a/NvCloth/src/SwClothData.cpp
+++ b/NvCloth/src/SwClothData.cpp
@@ -80,6 +80,7 @@ cloth::SwClothData::SwClothData(SwCloth& cloth, const SwFabric& fabric)
 	mNumTriangles = uint32_t(fabric.mTriangles.size()) / 3;
 	mDragCoefficient = 1.0f - expf(stiffnessExponent * cloth.mDragLogCoefficient);
 	mLiftCoefficient = 1.0f - expf(stiffnessExponent * cloth.mLiftLogCoefficient);
+	mFluidDensity = cloth.mFluidDensity * 0.5f; //divide by 2 to so we don't have to compensate for double area from cross product in the solver
 
 	mStartMotionConstraints = cloth.mMotionConstraints.mStart.size() ? array(cloth.mMotionConstraints.mStart.front()) : 0;
 	mTargetMotionConstraints =
diff --git a/NvCloth/src/SwClothData.h b/NvCloth/src/SwClothData.h
index d2387b5..78a6f99 100644
--- a/NvCloth/src/SwClothData.h
+++ b/NvCloth/src/SwClothData.h
@@ -92,6 +92,7 @@ struct SwClothData
 	uint32_t mNumTriangles;
 	float mDragCoefficient;
 	float mLiftCoefficient;
+	float mFluidDensity;
 
 	// motion constraint data
 	const float* mStartMotionConstraints;
diff --git a/NvCloth/src/SwSolverKernel.cpp b/NvCloth/src/SwSolverKernel.cpp
index 52dfdaa..dec46d7 100644
--- a/NvCloth/src/SwSolverKernel.cpp
+++ b/NvCloth/src/SwSolverKernel.cpp
@@ -384,9 +384,13 @@ T4f calculateMaxDelta(const T4f* prevIt, const T4f* curIt, const T4f* curEnd)
 
 template <bool IsTurning, typename T4f>
 void applyWind(T4f* __restrict curIt, const T4f* __restrict prevIt, const uint16_t* __restrict tIt,
-               const uint16_t* __restrict tEnd, T4f dragCoefficient, T4f liftCoefficient, T4f wind,
+               const uint16_t* __restrict tEnd, float itrDtf, float dragCoefficientf, float liftCoefficientf, float fluidDensityf, T4f wind,
                const T4f (&rotation)[3])
 {
+	const T4f dragCoefficient = simd4f(dragCoefficientf);
+	const T4f liftCoefficient = simd4f(liftCoefficientf);
+	const T4f fluidDensity = simd4f(fluidDensityf);
+	const T4f itrDt = simd4f(itrDtf);
 	const T4f oneThird = simd4f(1.0f / 3.0f);
 
 	for (; tIt < tEnd; tIt += 3)
@@ -410,7 +414,7 @@ void applyWind(T4f* __restrict curIt, const T4f* __restrict prevIt, const uint16
 		T4f previous = oneThird * (p0 + p1 + p2);
 
 		//offset of the triangle center, including wind
-		T4f delta = current - previous + wind;
+		T4f delta = current - previous + wind; //wind is also already multiplied by dt in the iteration state so everything it in the same units
 
 		if (IsTurning)
 		{
@@ -423,23 +427,25 @@ void applyWind(T4f* __restrict curIt, const T4f* __restrict prevIt, const uint16
 		T4f normal = cross3(c2 - c0, c1 - c0);
 
 		T4f doubleArea = sqrt(dot3(normal, normal));
+		normal = normal / doubleArea;
 
 		T4f invSqrScale = dot3(delta, delta);
 		T4f isZero = invSqrScale < gSimd4fEpsilon;
 		T4f scale = rsqrt(invSqrScale);
+		T4f deltaLength = sqrt(invSqrScale);
 
 		//scale 'normalizes' delta, doubleArea normalized normal
-		T4f cosTheta = dot3(normal, delta) * scale / doubleArea;
+		T4f cosTheta = dot3(normal, delta) * scale;
 		T4f sinTheta = sqrt(max(gSimd4fZero, gSimd4fOne - cosTheta * cosTheta));
 
 		// orthogonal to delta, in delta-normal plane, same length as delta
 		T4f liftDir = cross3(cross3(delta, normal), delta * scale);
 
 		// sin(theta) * cos(theta) = 0.5 * sin(2 * theta)
-		T4f lift = liftCoefficient * cosTheta * sinTheta * liftDir;
-		T4f drag = dragCoefficient * abs(cosTheta) * doubleArea * delta; //dragCoefficient should compensate for double area
+		T4f lift = liftCoefficient * cosTheta * sinTheta * liftDir * deltaLength / itrDt;
+		T4f drag = dragCoefficient * abs(cosTheta) * delta * deltaLength / itrDt; 
 
-		T4f impulse = (lift + drag) & ~isZero;
+		T4f impulse = (drag + lift) * fluidDensity * doubleArea & ~isZero; //fluidDensity compensates for double area
 
 		curIt[i0] = c0 - impulse * splat<3>(c0);
 		curIt[i1] = c1 - impulse * splat<3>(c1);
@@ -668,17 +674,14 @@ void cloth::SwSolverKernel<T4f>::applyWind()
 	const uint16_t* tIt = mClothData.mTriangles;
 	const uint16_t* tEnd = tIt + 3 * mClothData.mNumTriangles;
 
-	T4f dragCoefficient = simd4f(mClothData.mDragCoefficient);
-	T4f liftCoefficient = simd4f(mClothData.mLiftCoefficient);
-
 	if (mState.mIsTurning)
 	{
-		::applyWind<true>(curIt, prevIt, tIt, tEnd, dragCoefficient, liftCoefficient, mState.mWind,
+		::applyWind<true>(curIt, prevIt, tIt, tEnd, mState.mIterDt, mClothData.mDragCoefficient, mClothData.mLiftCoefficient, mClothData.mFluidDensity, mState.mWind,
 		                  mState.mRotationMatrix);
 	}
 	else
 	{
-		::applyWind<false>(curIt, prevIt, tIt, tEnd, dragCoefficient, liftCoefficient, mState.mWind,
+		::applyWind<false>(curIt, prevIt, tIt, tEnd, mState.mIterDt, mClothData.mDragCoefficient, mClothData.mLiftCoefficient, mClothData.mFluidDensity, mState.mWind,
 		                   mState.mRotationMatrix);
 	}
 }
diff --git a/NvCloth/src/cuda/CuClothData.cpp b/NvCloth/src/cuda/CuClothData.cpp
index decfd2c..927997c 100644
--- a/NvCloth/src/cuda/CuClothData.cpp
+++ b/NvCloth/src/cuda/CuClothData.cpp
@@ -124,6 +124,7 @@ cloth::CuFrameData::CuFrameData(CuCloth& cloth, uint32_t numSharedPositions, con
 	stiffness = gSimd4fOne - exp2(logStiffness * stiffnessExponent);
 	mDragCoefficient = array(stiffness)[0];
 	mLiftCoefficient = array(stiffness)[1];
+	mFluidDensity = cloth.mFluidDensity * 0.5f; //divide by 2 to so we don't have to compensate for double area from cross product in the solver
 	for (int i = 0; i < 9; ++i)
 		mRotation[i] = array(state.mRotationMatrix[i / 3])[i % 3];
 
diff --git a/NvCloth/src/cuda/CuClothData.h b/NvCloth/src/cuda/CuClothData.h
index 0e4cda0..dd836fd 100644
--- a/NvCloth/src/cuda/CuClothData.h
+++ b/NvCloth/src/cuda/CuClothData.h
@@ -135,6 +135,7 @@ struct CuFrameData
 	// wind data
 	float mDragCoefficient;
 	float mLiftCoefficient;
+	float mFluidDensity;
 	float mRotation[9];
 
 	// motion constraint data
diff --git a/NvCloth/src/cuda/CuSolverKernel.cu b/NvCloth/src/cuda/CuSolverKernel.cu
index 3517193..edb66dc 100644
--- a/NvCloth/src/cuda/CuSolverKernel.cu
+++ b/NvCloth/src/cuda/CuSolverKernel.cu
@@ -867,6 +867,8 @@ __device__ void applyWind(CurrentT& current, PreviousT& previous)
 {
 	const float dragCoefficient = gFrameData.mDragCoefficient;
 	const float liftCoefficient = gFrameData.mLiftCoefficient;
+	const float fluidDensity = gFrameData.mFluidDensity;
+	const float itrDt = gFrameData.mIterDt;
 
 	if (dragCoefficient == 0.0f && liftCoefficient == 0.0f)
 		return;
@@ -912,20 +914,22 @@ __device__ void applyWind(CurrentT& current, PreviousT& previous)
 
 		float3 normal = cross3(c2 - c0, c1 - c0);
 
-		float doubleArea = sqrtf(dot3(normal, normal));
+		const float doubleArea = sqrtf(dot3(normal, normal));
+		normal = (1.0f / doubleArea) * normal;
 
 		float invSqrScale = dot3(delta, delta);
 		float scale = rsqrtf(invSqrScale);
+		float deltaLength = sqrtf(invSqrScale);
 
-		float cosTheta = dot3(normal, delta) * scale / doubleArea;
+		float cosTheta = dot3(normal, delta) * scale;
 		float sinTheta = sqrtf(max(0.0f, 1.0f - cosTheta * cosTheta));
 
 		float3 liftDir = cross3(cross3(delta, normal), scale * delta);
 
-		float3 lift = liftCoefficient * cosTheta * sinTheta * liftDir;
-		float3 drag = dragCoefficient * abs(cosTheta) * doubleArea * delta;
+		float3 lift = liftCoefficient * cosTheta * sinTheta * ((deltaLength / itrDt) * liftDir);
+		float3 drag = dragCoefficient * abs(cosTheta) * ((deltaLength / itrDt) * delta);
 
-		float3 impulse = invSqrScale < FLT_EPSILON ? make_float3(0.0f, 0.0f, 0.0f) : lift + drag;
+		float3 impulse = invSqrScale < FLT_EPSILON ? make_float3(0.0f, 0.0f, 0.0f) : fluidDensity * doubleArea * (lift + drag);
 
 		applyImpulse(current(i0), impulse);
 		applyImpulse(current(i1), impulse);
diff --git a/NvCloth/src/dx/DxClothData.cpp b/NvCloth/src/dx/DxClothData.cpp
index 075dc81..57049bf 100644
--- a/NvCloth/src/dx/DxClothData.cpp
+++ b/NvCloth/src/dx/DxClothData.cpp
@@ -112,6 +112,7 @@ cloth::DxFrameData::DxFrameData(DxCloth& cloth, uint32_t numSharedPositions, con
 		Simd4f stiffness = gSimd4fOne - exp2(logStiffness * stiffnessExponent);
 		mDragCoefficient = array(stiffness)[0];
 		mLiftCoefficient = array(stiffness)[1];
+		mFluidDensity = cloth.mFluidDensity * 0.5f; //divide by 2 to so we don't have to compensate for double area from cross product in the solver
 		for(int i = 0; i < 9; ++i)
 			mRotation[i] = array(state.mRotationMatrix[i / 3])[i % 3];
 	}
diff --git a/NvCloth/src/dx/DxClothData.h b/NvCloth/src/dx/DxClothData.h
index af02bc6..f91d37d 100644
--- a/NvCloth/src/dx/DxClothData.h
+++ b/NvCloth/src/dx/DxClothData.h
@@ -46,7 +46,7 @@ typedef unsigned int uint32_t;
 typedef int int32_t;
 #endif
 
-static const uint32_t MaxParticlesInSharedMem = 1972;
+static const uint32_t MaxParticlesInSharedMem = 1971;
 
 
 struct DxPhaseConfig
@@ -167,6 +167,7 @@ struct DxFrameData
 	// wind data
 	float mDragCoefficient;
 	float mLiftCoefficient;
+	float mFluidDensity;
 	float mRotation[9];
 
 	// motion constraint data
diff --git a/NvCloth/src/dx/DxSolverKernel.hlsl b/NvCloth/src/dx/DxSolverKernel.hlsl
index 2ca42b3..4d91f4b 100644
--- a/NvCloth/src/dx/DxSolverKernel.hlsl
+++ b/NvCloth/src/dx/DxSolverKernel.hlsl
@@ -202,7 +202,8 @@ void applyWind(IParticles curParticles, IParticles prevParticles, uint32_t threa
 {
 	const float dragCoefficient = gFrameData.mDragCoefficient;
 	const float liftCoefficient = gFrameData.mLiftCoefficient;
-
+	const float fluidDensity = gFrameData.mFluidDensity;
+	const float itrDt = gFrameData.mIterDt;
 	if(dragCoefficient == 0.0f && liftCoefficient == 0.0f)
 		return;
 
@@ -258,20 +259,22 @@ void applyWind(IParticles curParticles, IParticles prevParticles, uint32_t threa
 
 		float3 normal = cross(c2.xyz - c0.xyz, c1.xyz - c0.xyz);
 
-		float doubleArea = sqrt(dot(normal, normal));
+		const float doubleArea = sqrt(dot(normal, normal));
+		normal = normal / doubleArea;
 
 		float invSqrScale = dot(delta, delta);
 		float scale = rsqrt(invSqrScale);
+		float deltaLength = sqrt(invSqrScale);
 
-		float cosTheta = dot(normal, delta) * scale / doubleArea;
+		float cosTheta = dot(normal, delta) * scale;
 		float sinTheta = sqrt(max(0.0f, 1.0f - cosTheta * cosTheta));
 
 		float3 liftDir = cross(cross(delta, normal), scale * delta);
 
-		float3 lift = liftCoefficient * cosTheta * sinTheta * liftDir;
-		float3 drag = dragCoefficient * abs(cosTheta) * doubleArea * delta;
+		float3 lift = liftCoefficient * cosTheta * sinTheta * liftDir * deltaLength / itrDt;
+		float3 drag = dragCoefficient * abs(cosTheta) * delta * deltaLength / itrDt;
 
-		float3 impulse = invSqrScale < 1.192092896e-07F ? float3(0.0f, 0.0f, 0.0f) : lift + drag;
+		float3 impulse = invSqrScale < 1.192092896e-07F ? float3(0.0f, 0.0f, 0.0f) : (lift + drag) * fluidDensity * doubleArea;
 
 		curParticles.atomicAdd(i0, -impulse * c0.w);
 		curParticles.atomicAdd(i1, -impulse * c1.w);
diff --git a/NvCloth/src/neon/SwCollisionHelpers.h b/NvCloth/src/neon/SwCollisionHelpers.h
index 0b9410b..acd8a2a 100644
--- a/NvCloth/src/neon/SwCollisionHelpers.h
+++ b/NvCloth/src/neon/SwCollisionHelpers.h
@@ -33,7 +33,7 @@
 #include <arm_neon.h>
 #endif
 
-namespace physx
+namespace nv
 {
 namespace cloth
 {
@@ -84,4 +84,4 @@ Simd4i Gather<Simd4i>::operator()(const Simd4i* ptr) const
 }
 
 } // namespace cloth
-} // namespace physx
+} // namespace nv
diff --git a/NvCloth/src/scalar/SwCollisionHelpers.h b/NvCloth/src/scalar/SwCollisionHelpers.h
index 0d7321f..af21812 100644
--- a/NvCloth/src/scalar/SwCollisionHelpers.h
+++ b/NvCloth/src/scalar/SwCollisionHelpers.h
@@ -29,7 +29,7 @@
 
 #pragma once
 
-namespace physx
+namespace nv
 {
 namespace cloth
 {
@@ -89,4 +89,4 @@ Scalar4i Gather<Scalar4i>::operator()(const Scalar4i* ptr) const
 }
 
 } // namespace cloth
-} // namespace physx
+} // namespace nv
diff --git a/PxShared/include/cudamanager/PxCudaContextManager.h b/PxShared/include/cudamanager/PxCudaContextManager.h
index 24f4370..aca1112 100644
--- a/PxShared/include/cudamanager/PxCudaContextManager.h
+++ b/PxShared/include/cudamanager/PxCudaContextManager.h
@@ -23,7 +23,7 @@
 // components in life support devices or systems without express written approval of
 // NVIDIA Corporation.
 //
-// Copyright (c) 2008-2014 NVIDIA Corporation. All rights reserved.
+// Copyright (c) 2008-2017 NVIDIA Corporation. All rights reserved.
 
 
 #ifndef PXCUDACONTEXTMANAGER_PXCUDACONTEXTMANAGER_H
@@ -42,6 +42,7 @@
 /* Forward decl to avoid inclusion of cuda.h */
 typedef struct CUctx_st *CUcontext;
 typedef struct CUgraphicsResource_st *CUgraphicsResource;
+typedef int CUdevice;
 
 namespace physx
 { 
@@ -58,7 +59,6 @@ struct PxCudaInteropMode
 	enum Enum
 	{
 		NO_INTEROP = 0,
-		D3D9_INTEROP,
 		D3D10_INTEROP,
 		D3D11_INTEROP,
 		OGL_INTEROP,
@@ -154,6 +154,9 @@ public:
      * to every CUDA work submission, so we recommend that you carefully tune
      * this initial base memory size to closely approximate the amount of
      * memory your application will consume.
+
+	 Note: This is currently not used by PxSceneFlag::eENABLE_GPU_DYNAMICS. Memory allocation properties are configured
+	 for GPU rigid bodies using PxSceneDesc::gpuDynamicsConfig.
      */
 	uint32_t	memoryBaseSize[PxCudaBufferMemorySpace::COUNT];
 
@@ -162,11 +165,17 @@ public:
      *
      * The memory manager will dynamically grow and shrink in blocks multiple of
      * this page size. Size has to be power of two and bigger than 0.
+
+	Note: This is currently not used by PxSceneFlag::eENABLE_GPU_DYNAMICS. Memory allocation properties are configured
+	for GPU rigid bodies using PxSceneDesc::gpuDynamicsConfig.
      */
 	uint32_t	memoryPageSize[PxCudaBufferMemorySpace::COUNT];
 
     /**
      * \brief Maximum size of memory that the memory manager will allocate
+
+	 Note: This is currently not used by PxSceneFlag::eENABLE_GPU_DYNAMICS. Memory allocation properties are configured
+	 for GPU rigid bodies using PxSceneDesc::gpuDynamicsConfig.
      */
 	uint32_t	maxMemorySize[PxCudaBufferMemorySpace::COUNT];
 
@@ -223,7 +232,7 @@ public:
      * harmfull to (re)acquire the context in code that is shared between
      * GpuTasks and non-task functions.
      */
-	virtual CUcontext acquireContext() = 0;
+    virtual void acquireContext() = 0;
 
     /**
      * \brief Release the CUDA context from the current thread
@@ -234,9 +243,16 @@ public:
      */
     virtual void releaseContext() = 0;
 
+	/**
+	* \brief Return the CUcontext
+	*/
+	virtual CUcontext getContext() = 0;
+
     /**
      * \brief Return the PxCudaMemoryManager instance associated with this
      * CUDA context
+	 * Note: This is currently not used by PxSceneFlag::eENABLE_GPU_DYNAMICS. Memory allocation properties are configured
+	 * for GPU rigid bodies using PxSceneDesc::gpuDynamicsConfig.
      */
 	virtual PxCudaMemoryManager *getMemoryManager() = 0;
 
@@ -268,6 +284,7 @@ public:
 	virtual bool supportsArchSM35() const = 0;  //!< GK110
 	virtual bool supportsArchSM50() const = 0;  //!< GM100
 	virtual bool supportsArchSM52() const = 0;  //!< GM200
+	virtual bool supportsArchSM60() const = 0;  //!< GP100
 	virtual bool isIntegrated() const = 0;      //!< true if GPU is an integrated (MCP) part
 	virtual bool canMapHostMemory() const = 0;  //!< true if GPU map host memory to GPU (0-copy)
 	virtual int  getDriverVersion() const = 0;  //!< returns cached value of cuGetDriverVersion()
@@ -278,6 +295,7 @@ public:
 	virtual int  getSharedMemPerMultiprocessor() const = 0; //!< returns total amount of shared memory available per multiprocessor in bytes
 	virtual unsigned int getMaxThreadsPerBlock() const = 0; //!< returns the maximum number of threads per block
     virtual const char *getDeviceName() const = 0; //!< returns device name retrieved from driver
+	virtual CUdevice getDevice() const = 0; //!< returns device handle retrieved from driver
 	virtual PxCudaInteropMode::Enum getInteropMode() const = 0; //!< interop mode the context was created with
 
 	virtual void setUsingConcurrentStreams(bool) = 0; //!< turn on/off using concurrent streams for GPU work
diff --git a/PxShared/include/foundation/PxIntrinsics.h b/PxShared/include/foundation/PxIntrinsics.h
index 471f934..b4aff28 100644
--- a/PxShared/include/foundation/PxIntrinsics.h
+++ b/PxShared/include/foundation/PxIntrinsics.h
@@ -38,8 +38,8 @@
 #include "foundation/unix/PxUnixIntrinsics.h"
 #elif PX_XBOXONE
 #include "foundation/XboxOne/PxXboxOneIntrinsics.h"
-#elif PX_NX
-#include "foundation/nx/PxNXIntrinsics.h"
+#elif PX_SWITCH
+#include "foundation/switch/PxSwitchIntrinsics.h"
 #else
 #error "Platform not supported!"
 #endif
diff --git a/PxShared/include/foundation/PxPreprocessor.h b/PxShared/include/foundation/PxPreprocessor.h
index 446ca76..9b6e0f4 100644
--- a/PxShared/include/foundation/PxPreprocessor.h
+++ b/PxShared/include/foundation/PxPreprocessor.h
@@ -88,7 +88,7 @@ Operating system defines, see http://sourceforge.net/p/predef/wiki/OperatingSyst
 #elif defined(__ORBIS__)
 #define PX_PS4 1
 #elif defined(__NX__)
-#define PX_NX 1
+#define PX_SWITCH 1
 #else
 #error "Unknown operating system"
 #endif
@@ -161,8 +161,8 @@ define anything not defined on this platform to 0
 #ifndef PX_PS4
 #define PX_PS4 0
 #endif
-#ifndef PX_NX
-#define PX_NX 0
+#ifndef PX_SWITCH
+#define PX_SWITCH 0
 #endif
 #ifndef PX_X64
 #define PX_X64 0
@@ -425,7 +425,7 @@ General defines
 */
 
 // static assert
-#if(defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 7))) || (PX_PS4) || (PX_APPLE_FAMILY) || (PX_NX) || (PX_CLANG && PX_ARM)
+#if(defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 7))) || (PX_PS4) || (PX_APPLE_FAMILY) || (PX_SWITCH) || (PX_CLANG && PX_ARM)
 #define PX_COMPILE_TIME_ASSERT(exp) typedef char PxCompileTimeAssert_Dummy[(exp) ? 1 : -1] __attribute__((unused))
 #else
 #define PX_COMPILE_TIME_ASSERT(exp) typedef char PxCompileTimeAssert_Dummy[(exp) ? 1 : -1]
@@ -532,5 +532,11 @@ protected:
 
 #define PX_SUPPORT_COMPUTE_PHYSX 0
 
+#ifndef PX_SUPPORT_EXTERN_TEMPLATE
+#define PX_SUPPORT_EXTERN_TEMPLATE ((!PX_ANDROID) && (PX_VC != 11))
+#else
+#define PX_SUPPORT_EXTERN_TEMPLATE 0
+#endif
+
 /** @} */
 #endif // #ifndef PXFOUNDATION_PXPREPROCESSOR_H
diff --git a/PxShared/include/foundation/nx/PxNXIntrinsics.h b/PxShared/include/foundation/nx/PxNXIntrinsics.h
deleted file mode 100644
index adf3bf5..0000000
--- a/PxShared/include/foundation/nx/PxNXIntrinsics.h
+++ /dev/null
@@ -1,138 +0,0 @@
-/*
- * Copyright (c) 2008-2015, NVIDIA CORPORATION.  All rights reserved.
- *
- * NVIDIA CORPORATION and its licensors retain all intellectual property
- * and proprietary rights in and to this software, related documentation
- * and any modifications thereto.  Any use, reproduction, disclosure or
- * distribution of this software and related documentation without an express
- * license agreement from NVIDIA CORPORATION is strictly prohibited.
- */
-// Copyright (c) 2004-2008 AGEIA Technologies, Inc. All rights reserved.
-// Copyright (c) 2001-2004 NovodeX AG. All rights reserved.  
-
-
-#ifndef PX_FOUNDATION_PX_NX_INTRINSICS_H
-#define PX_FOUNDATION_PX_NX_INTRINSICS_H
-
-#include "foundation/Px.h"
-#include "foundation/PxAssert.h"
-
-#if !PX_NX
-	#error "This file should only be included by NX builds!!"
-#endif
-
-#include <math.h>
-#include <float.h>
-
-#include "nn/cstd/cstd_CMath.h"
-
-#if !PX_DOXYGEN
-namespace physx
-{
-namespace intrinsics
-{
-#endif
-
-	//! \brief platform-specific absolute value
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float abs(float a)						{	return ::fabsf(a);	}
-
-	//! \brief platform-specific select float
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float fsel(float a, float b, float c)	{	return (a >= 0.0f) ? b : c;	}
-
-	//! \brief platform-specific sign
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float sign(float a)					{	return (a >= 0.0f) ? 1.0f : -1.0f; }
-
-	//! \brief platform-specific reciprocal
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float recip(float a)					{	return 1.0f/a;			}
-
-	//! \brief platform-specific reciprocal estimate
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipFast(float a)				{	return 1.0f/a;			}
-
-	//! \brief platform-specific square root
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float sqrt(float a)					{	return ::sqrtf(a);	}
-
-	//! \brief platform-specific reciprocal square root
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipSqrt(float a)				{   return 1.0f/::sqrtf(a); }
-
-	//! \brief platform-specific reciprocal square root estimate
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipSqrtFast(float a)			{	return 1.0f/::sqrtf(a); }
-
-	//! \brief platform-specific sine
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float sin(float a)						{   return ::sinf(a); }
-
-	//! \brief platform-specific cosine
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float cos(float a)						{   return ::cosf(a); }
-
-	//! \brief platform-specific minimum
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float selectMin(float a, float b)		{	return a<b ? a : b;	}
-
-	//! \brief platform-specific maximum
-	PX_CUDA_CALLABLE PX_FORCE_INLINE float selectMax(float a, float b)		{	return a>b ? a : b; }
-
-	//! \brief platform-specific finiteness check
-	PX_CUDA_CALLABLE PX_FORCE_INLINE bool isFinite(float a)
-	{
-#ifdef __CUDACC__
-		return isfinite(a) ? true : false;
-#else
-		return !nn::cstd::IsNan(a) && !nn::cstd::IsInf(a);
-#endif
-	}
-
-	//! \brief platform-specific finiteness check
-	PX_CUDA_CALLABLE PX_FORCE_INLINE bool isFinite(double a)
-	{
-#ifdef __CUDACC__
-		return isfinite(a) ? true : false;
-#else
-		return !nn::cstd::IsNan(a) && !nn::cstd::IsInf(a);
-#endif
-	}
-
-	/*!
-	Sets \c count bytes starting at \c dst to zero.
-	*/
-	PX_FORCE_INLINE void* memZero(void* PX_RESTRICT dest, uint32_t count)
-	{
-		return memset(dest, 0, count);
-	}
-
-	/*!
-	Sets \c count bytes starting at \c dst to \c c.
-	*/
-	PX_FORCE_INLINE void* memSet(void* PX_RESTRICT dest, int32_t c, uint32_t count)
-	{
-		return memset(dest, c, count);
-	}
-
-	/*!
-	Copies \c count bytes from \c src to \c dst. User memMove if regions overlap.
-	*/
-	PX_FORCE_INLINE void* memCopy(void* PX_RESTRICT dest, const void* PX_RESTRICT src, uint32_t count)
-	{
-		return memcpy(dest, src, count);
-	}
-
-	/*!
-	Copies \c count bytes from \c src to \c dst. Supports overlapping regions.
-	*/
-	PX_FORCE_INLINE void* memMove(void* PX_RESTRICT dest, const void* PX_RESTRICT src, uint32_t count)
-	{
-		return memmove(dest, src, count);
-	}
-
-	/*!
-	Set 128B to zero starting at \c dst+offset. Must be aligned.
-	*/
-	PX_FORCE_INLINE void memZero128(void* PX_RESTRICT dest, uint32_t offset = 0)
-	{
-		PX_ASSERT(((size_t(dest)+offset) & 0x7f) == 0);
-		memSet((char* PX_RESTRICT)dest+offset, 0, 128);
-	}
-
-#if !PX_DOXYGEN
-} // namespace intrinsics
-} // namespace physx
-#endif
-
-#endif
diff --git a/PxShared/include/task/PxTask.h b/PxShared/include/task/PxTask.h
index 2761109..85d91da 100644
--- a/PxShared/include/task/PxTask.h
+++ b/PxShared/include/task/PxTask.h
@@ -45,7 +45,7 @@ namespace physx
 class PxBaseTask
 {
 public:
-	PxBaseTask() : mEventID(0xFFFF), mProfileStat(0), mTm(0) {}
+	PxBaseTask() : mContextID(0), mTm(NULL) {}
 	virtual ~PxBaseTask() {}
 
     /**
@@ -78,50 +78,24 @@ public:
 	 * references to it - so it may safely run its destructor, recycle itself, etc.
 	 * provided no additional user references to the task exist
 	 */
-
     virtual void		release() = 0;
 
-	/**
-     * \brief Execute user run method with wrapping profiling events.
-     *
-     * Optional entry point for use by CpuDispatchers.
-	 *
-	 * \param[in] threadId The threadId of the thread that executed the task.
-	 */
-	PX_INLINE void runProfiled(uint32_t threadId=0)
-	{		
-		mTm->emitStartEvent(*this, threadId);
-		run();
-		mTm->emitStopEvent(*this, threadId);
-	}
-
-	/**
-     * \brief Specify stop event statistic
-     *
-     * If called before or while the task is executing, the given value
-	 * will appear in the task's event bar in the profile viewer
-	 *
-	 * \param[in] stat The stat to signal when the task is finished
-	 */
-	PX_INLINE void setProfileStat( uint16_t stat )
-	{
-		mProfileStat = stat;
-	}
-
     /**
      * \brief Return PxTaskManager to which this task was submitted
      *
      * Note, can return NULL if task was not submitted, or has been
      * completed.
      */
-	PX_INLINE PxTaskManager* getTaskManager() const
+	PX_FORCE_INLINE PxTaskManager* getTaskManager() const
 	{
 		return mTm;
 	}
 
+	PX_FORCE_INLINE	void	setContextId(PxU64 id)			{ mContextID = id;		}
+	PX_FORCE_INLINE	PxU64	getContextId()			const	{ return mContextID;	}
+
 protected:
-	uint16_t			mEventID;		//!< Registered profile event ID
-	uint16_t			mProfileStat;	//!< Profiling statistic
+	PxU64				mContextID;		//!< Context ID for profiler interface
 	PxTaskManager*		mTm;			//!< Owning PxTaskManager instance
 
 	friend class PxTaskMgr;
@@ -212,7 +186,6 @@ public:
 	{
 		mStreamIndex = 0;
 		mPreSyncRequired = false;
-		mProfileStat = 0;
 	}
 
 	/**
diff --git a/PxShared/include/task/PxTaskManager.h b/PxShared/include/task/PxTaskManager.h
index f6f29c4..f40f7b1 100644
--- a/PxShared/include/task/PxTaskManager.h
+++ b/PxShared/include/task/PxTaskManager.h
@@ -215,9 +215,6 @@ protected:
 	virtual void decrReference(PxLightCpuTask&) = 0;
 	virtual void addReference(PxLightCpuTask&) = 0;
 
-	virtual void emitStartEvent(PxBaseTask&, uint32_t threadId=0) = 0;
-	virtual void emitStopEvent(PxBaseTask&, uint32_t threadId=0) = 0;
-
 	/*! \endcond */
 
 	friend class PxBaseTask;
diff --git a/PxShared/src/compiler/cmake/Android/CMakeLists.txt b/PxShared/src/compiler/cmake/Android/CMakeLists.txt
index 0499c29..9aa00f9 100644
--- a/PxShared/src/compiler/cmake/Android/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/Android/CMakeLists.txt
@@ -1,34 +1,37 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared )
 include(../common/CMakeLists.txt)
 
 STRING(TOLOWER "${CMAKE_BUILD_TYPE}" CMAKE_BUILD_TYPE_LOWERCASE)
 
-IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Android
-	SET(TARGET_BUILD_PLATFORM "Android")
+IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to android
+	SET(TARGET_BUILD_PLATFORM "android")
 ENDIF()
 
-SET(PLATFORM_LIST Android)
+SET(PLATFORM_LIST android)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
 ENDIF()
 
 if(${ANDROID_ABI} STREQUAL "armeabi-v7a")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -D__STDC_LIMIT_MACROS -Wno-invalid-offsetof ")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffunction-sections -fdata-sections -funwind-tables -fomit-frame-pointer -funswitch-loops -finline-limit=300 -fno-strict-aliasing -fstack-protector -Wno-invalid-offsetof ")
+elseif(${ANDROID_ABI} STREQUAL "armeabi-v7a with NEON")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -fno-rtti -ffast-math -fno-exceptions -ffunction-sections -funwind-tables -fomit-frame-pointer -funswitch-loops -finline-limit=300 -fno-strict-aliasing -fstack-protector -Wno-invalid-offsetof ")
 elseif(${ANDROID_ABI} STREQUAL "arm64-v8a")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -D__STDC_LIMIT_MACROS -Wno-invalid-offsetof ")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffunction-sections -fdata-sections -Wno-invalid-offsetof ")
 elseif(${ANDROID_ABI} STREQUAL "x86")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -D__STDC_LIMIT_MACROS -Wno-invalid-offsetof -fpack-struct=8 -malign-double ")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffunction-sections -fdata-sections -Wno-invalid-offsetof -fpack-struct=8 -malign-double ")
 elseif(${ANDROID_ABI} STREQUAL "x86_64")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -D__STDC_LIMIT_MACROS -Wno-invalid-offsetof -mstackrealign -msse3  ")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS}  -std=c++11 -fno-rtti -fno-exceptions -ffunction-sections -fdata-sections -Wno-invalid-offsetof -mstackrealign -msse3  ")
 endif()
 	
 SET(CMAKE_CXX_FLAGS_DEBUG "-O0 -g")
-SET(CMAKE_CXX_FLAGS_CHECKED "-O2")
-SET(CMAKE_CXX_FLAGS_PROFILE "-O2")
-SET(CMAKE_CXX_FLAGS_RELEASE "-O2")
+SET(CMAKE_CXX_FLAGS_CHECKED "-O3")
+SET(CMAKE_CXX_FLAGS_PROFILE "-O3")
+SET(CMAKE_CXX_FLAGS_RELEASE "-O3")
 
-SET(PXSHARED_ANDROID_COMPILE_DEFS _LIB;)
+SET(PXSHARED_ANDROID_COMPILE_DEFS _LIB;__STDC_LIMIT_MACROS;)
 SET(PXSHARED_ANDROID_DEBUG_COMPILE_DEFS _DEBUG;PX_DEBUG=1;PX_CHECKED=1)
 SET(PXSHARED_ANDROID_CHECKED_COMPILE_DEFS NDEBUG;PX_CHECKED=1)
 SET(PXSHARED_ANDROID_PROFILE_COMPILE_DEFS NDEBUG;PX_PROFILE=1)
diff --git a/PxShared/src/compiler/cmake/IOS/CMakeLists.txt b/PxShared/src/compiler/cmake/IOS/CMakeLists.txt
index d281e32..4e2f814 100644
--- a/PxShared/src/compiler/cmake/IOS/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/IOS/CMakeLists.txt
@@ -1,12 +1,13 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared CXX)
 include(../common/CMakeLists.txt)
 
 
-IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to IOS
-	SET(TARGET_BUILD_PLATFORM "IOS")
+IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to ios
+	SET(TARGET_BUILD_PLATFORM "ios")
 ENDIF()
 
-SET(PLATFORM_LIST IOS)
+SET(PLATFORM_LIST ios)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
diff --git a/PxShared/src/compiler/cmake/Linux/CMakeLists.txt b/PxShared/src/compiler/cmake/Linux/CMakeLists.txt
index 2fa592e..ea25a8b 100644
--- a/PxShared/src/compiler/cmake/Linux/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/Linux/CMakeLists.txt
@@ -1,13 +1,14 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared CXX)
 include(../common/CMakeLists.txt)
 
 STRING(TOLOWER "${CMAKE_BUILD_TYPE}" CMAKE_BUILD_TYPE_LOWERCASE)
 
-IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Linux
-	SET(TARGET_BUILD_PLATFORM "Linux")
+IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to linux
+	SET(TARGET_BUILD_PLATFORM "linux")
 ENDIF()
 
-SET(PLATFORM_LIST Linux)
+SET(PLATFORM_LIST linux)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
@@ -16,16 +17,16 @@ ENDIF()
 IF (${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-unknown-linux-gnu" OR ${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-linux-gnu")
 	IF ("${CMAKE_CXX_COMPILER_ID}" STREQUAL "Clang")
 	  # using Clang	  
-	  SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-invalid-noreturn -Wno-return-type-c-linkage -Wno-reserved-id-macro -Wno-c++98-compat-pedantic -Wno-unused-local-typedef -Wno-old-style-cast -Wno-newline-eof -Wno-unused-private-field -Wno-undefined-func-template -Wno-format-nonliteral -Wno-implicit-fallthrough -Wno-undefined-reinterpret-cast")
+	  SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-invalid-noreturn -Wno-return-type-c-linkage -Wno-reserved-id-macro -Wno-c++98-compat-pedantic -Wno-unused-local-typedef -Wno-old-style-cast -Wno-newline-eof -Wno-unused-private-field -Wno-undefined-func-template -Wno-format-nonliteral -Wno-implicit-fallthrough -Wno-undefined-reinterpret-cast -Wno-disabled-macro-expansion")
 	ELSEIF ("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
 	  # using GCC
 	  SET(LIBPATH_SUFFIX "x64")
 	  SET(CMAKE_CXX_FLAGS "-Werror -m64 -fPIC -msse2 -mfpmath=sse -ffast-math -fno-exceptions -fno-rtti -fvisibility=hidden -fvisibility-inlines-hidden -Wall -Wextra -fno-strict-aliasing -fdiagnostics-show-option -Wno-invalid-offsetof -Wno-uninitialized -Wno-missing-field-initializers")	  
 	ENDIF("${CMAKE_CXX_COMPILER_ID}" STREQUAL "Clang")	
 ELSEIF(${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "arm-unknown-linux-gnueabihf")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -mfpu=neon -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-old-style-cast -Wno-return-type-c-linkage -Wno-format-nonliteral -Wno-implicit-fallthrough")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -mfpu=neon -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-old-style-cast -Wno-return-type-c-linkage -Wno-format-nonliteral -Wno-implicit-fallthrough -Wno-disabled-macro-expansion")
 ELSEIF(${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "aarch64-unknown-linux-gnueabi")
-	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-old-style-cast -Wno-return-type-c-linkage -Wno-format-nonliteral -Wno-unused-local-typedef -Wno-implicit-fallthrough")
+	SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -fno-rtti -fno-exceptions -ffast-math -ffunction-sections -fdata-sections -Werror -ferror-limit=0 -Wall -Wextra -fstrict-aliasing -Wstrict-aliasing=2 -Weverything -Wno-documentation-deprecated-sync -Wno-documentation-unknown-command -Wno-float-equal -Wno-padded -Wno-weak-vtables -Wno-cast-align -Wno-conversion -Wno-missing-noreturn -Wno-missing-variable-declarations -Wno-shift-sign-overflow -Wno-covered-switch-default -Wno-exit-time-destructors -Wno-global-constructors -Wno-missing-prototypes -Wno-unreachable-code -Wno-unused-macros -Wno-unused-member-function -Wno-used-but-marked-unused -Wno-weak-template-vtables -Wno-deprecated -Wno-non-virtual-dtor -Wno-old-style-cast -Wno-return-type-c-linkage -Wno-format-nonliteral -Wno-unused-local-typedef -Wno-implicit-fallthrough -Wno-disabled-macro-expansion")
 ELSE(${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-unknown-linux-gnu" OR ${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-linux-gnu")
 	MESSAGE(FATAL_ERROR "Unknown CMAKE_LIBRARY_ARCHITECTURE ${CMAKE_LIBRARY_ARCHITECTURE}")
 ENDIF(${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-unknown-linux-gnu" OR ${CMAKE_LIBRARY_ARCHITECTURE} STREQUAL "x86_64-linux-gnu")
@@ -36,7 +37,7 @@ SET(CMAKE_CXX_FLAGS_CHECKED "-O3 -g -gdwarf-3")
 SET(CMAKE_CXX_FLAGS_PROFILE "-O3 -g -gdwarf-3")
 SET(CMAKE_CXX_FLAGS_RELEASE "-O3 -g -gdwarf-3")
 
-IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+IF(DEFINED GENERATE_GPU_PROJECTS)
 SET(PXSHARED_LINUX_COMPILE_DEFS _LIB)
 ELSE()
 # Disable cuda and dx for all projects on windows
@@ -71,7 +72,7 @@ IF(DEFINED PX_SELECT_COMPONENTS)
 		INCLUDE(PxTask.cmake)
 	endif()
 	if ("PxCudaContextManager" IN_LIST PX_SELECT_COMPONENTS)
-		IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+		IF(DEFINED GENERATE_GPU_PROJECTS)
 			INCLUDE(PxCudaContextManager.cmake)
 		ENDIF()
 	endif()
@@ -80,7 +81,7 @@ INCLUDE(PxFoundation.cmake)
 INCLUDE(PsFastXml.cmake)
 INCLUDE(PxPvdSDK.cmake)
 INCLUDE(PxTask.cmake)
-IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+IF(DEFINED GENERATE_GPU_PROJECTS)
 	INCLUDE(PxCudaContextManager.cmake)
 ENDIF()
 ENDIF()
diff --git a/PxShared/src/compiler/cmake/Mac/CMakeLists.txt b/PxShared/src/compiler/cmake/Mac/CMakeLists.txt
index beb06bc..3242b46 100644
--- a/PxShared/src/compiler/cmake/Mac/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/Mac/CMakeLists.txt
@@ -1,12 +1,13 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared CXX)
 include(../common/CMakeLists.txt)
 
 
-IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Mac
-	SET(TARGET_BUILD_PLATFORM "Mac")
+IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to mac
+	SET(TARGET_BUILD_PLATFORM "mac")
 ENDIF()
 
-SET(PLATFORM_LIST Mac)
+SET(PLATFORM_LIST mac)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
diff --git a/PxShared/src/compiler/cmake/Modules/FindnvToolsExt.cmake b/PxShared/src/compiler/cmake/Modules/FindnvToolsExt.cmake
index c1675ae..f7bc761 100644
--- a/PxShared/src/compiler/cmake/Modules/FindnvToolsExt.cmake
+++ b/PxShared/src/compiler/cmake/Modules/FindnvToolsExt.cmake
@@ -13,7 +13,7 @@ FIND_PATH(		NVTOOLSEXT_INCLUDE_DIRS nvToolsExt.h
 
 INCLUDE(FindPackageHandleStandardArgs)
 				
-IF(TARGET_BUILD_PLATFORM STREQUAL "Windows")
+IF(TARGET_BUILD_PLATFORM STREQUAL "windows")
 	# NOTE: Doesn't make sense for all platforms - ARM
 	IF(CMAKE_CL_64)
 		SET(NVTOOLSEXT_LIBNAME "nvToolsExt64_1")
diff --git a/PxShared/src/compiler/cmake/common/CMakeLists.txt b/PxShared/src/compiler/cmake/common/CMakeLists.txt
index 22d2097..3bbb57a 100644
--- a/PxShared/src/compiler/cmake/common/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/common/CMakeLists.txt
@@ -1,7 +1,5 @@
 cmake_minimum_required(VERSION 3.3)
 
-PROJECT(PxShared CXX)
-
 CMAKE_POLICY(SET CMP0057 NEW) # Enable IN_LIST
 
 IF(DEFINED ENV{GW_DEPS_ROOT})
diff --git a/PxShared/src/compiler/cmake/html5/CMakeLists.txt b/PxShared/src/compiler/cmake/html5/CMakeLists.txt
index 8b9587a..e0d2776 100644
--- a/PxShared/src/compiler/cmake/html5/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/html5/CMakeLists.txt
@@ -1,10 +1,11 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared CXX)
 #set(CMAKE_VERBOSE_MAKEFILE ON)
 include(../common/CMakeLists.txt)
 
 
 IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to HTML5
-	SET(TARGET_BUILD_PLATFORM "HTML5")
+	SET(TARGET_BUILD_PLATFORM "html5")
 ENDIF()
 
 SET(PLATFORM_LIST HTML5)
diff --git a/PxShared/src/compiler/cmake/windows/CMakeLists.txt b/PxShared/src/compiler/cmake/windows/CMakeLists.txt
index 39b7dfc..fe2b00d 100644
--- a/PxShared/src/compiler/cmake/windows/CMakeLists.txt
+++ b/PxShared/src/compiler/cmake/windows/CMakeLists.txt
@@ -1,12 +1,13 @@
 cmake_minimum_required(VERSION 3.3)
+PROJECT(PxShared CXX)
 include(../common/CMakeLists.txt)
 
 
-IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to Windows
-	SET(TARGET_BUILD_PLATFORM "Windows")
+IF(NOT DEFINED TARGET_BUILD_PLATFORM) # Not defined, default to windows
+	SET(TARGET_BUILD_PLATFORM "windows")
 ENDIF()
 
-SET(PLATFORM_LIST Windows)
+SET(PLATFORM_LIST windows)
 
 IF (NOT ${TARGET_BUILD_PLATFORM} IN_LIST PLATFORM_LIST)
 	MESSAGE(FATAL_ERROR "Invalid platform:" ${TARGET_BUILD_PLATFORM})
@@ -33,7 +34,7 @@ SET(CMAKE_SHARED_LINKER_FLAGS "/DEBUG")
 # Controls PX_NVTX for all projects on windows
 SET(PXSHARED_WINDOWS_ENABLE_NVTX 0)	
 
-IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+IF(DEFINED LINK_GPU_BINARIES)
 SET(PXSHARED_WINDOWS_COMPILE_DEFS WIN32;_CRT_SECURE_NO_DEPRECATE;_CRT_NONSTDC_NO_DEPRECATE;_WINSOCK_DEPRECATED_NO_WARNINGS;)
 ELSE()
 # Disable cuda and dx for all projects on windows
@@ -44,6 +45,10 @@ SET(PXSHARED_WINDOWS_CHECKED_COMPILE_DEFS NDEBUG;PX_CHECKED=1;PX_NVTX=${PXSHARED
 SET(PXSHARED_WINDOWS_PROFILE_COMPILE_DEFS NDEBUG;PX_PROFILE=1;PX_NVTX=${PXSHARED_WINDOWS_ENABLE_NVTX})
 SET(PXSHARED_WINDOWS_RELEASE_COMPILE_DEFS NDEBUG)
 
+IF(DEFINED PX_SCALAR_MATH)
+	ADD_DEFINITIONS(-DPX_SIMD_DISABLED)
+ENDIF()
+
 IF(CMAKE_CL_64)
 	ADD_DEFINITIONS(-DWIN64)
 ENDIF(CMAKE_CL_64)		
@@ -74,7 +79,7 @@ IF(DEFINED PX_SELECT_COMPONENTS)
 		INCLUDE(PxTask.cmake)
 	endif()
 	if ("PxCudaContextManager" IN_LIST PX_SELECT_COMPONENTS)
-		IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+		IF(DEFINED GENERATE_GPU_PROJECTS)
 			INCLUDE(PxCudaContextManager.cmake)
 		ENDIF()
 	endif()
@@ -83,7 +88,7 @@ INCLUDE(PxFoundation.cmake)
 INCLUDE(PsFastXml.cmake)
 INCLUDE(PxPvdSDK.cmake)
 INCLUDE(PxTask.cmake)
-IF(DEFINED PX_GENERATE_GPU_PROJECTS)
+IF(DEFINED GENERATE_GPU_PROJECTS)
 	INCLUDE(PxCudaContextManager.cmake)
 ENDIF()
 ENDIF()
diff --git a/PxShared/src/cudamanager/include/GpuDispatcher.h b/PxShared/src/cudamanager/include/GpuDispatcher.h
index aedb345..10c412f 100644
--- a/PxShared/src/cudamanager/include/GpuDispatcher.h
+++ b/PxShared/src/cudamanager/include/GpuDispatcher.h
@@ -258,8 +258,6 @@ public:
 	~PxGpuWorkerThread();
 
 	void					setCudaContext(PxCudaContextManager& ctx);
-	void					emitStartEvent(const char *id);
-	void					emitStopEvent(const char *id);
 
 	/* API to TaskManager */
 	void					startSimulation();
diff --git a/PxShared/src/cudamanager/src/BlockingWait.cpp b/PxShared/src/cudamanager/src/BlockingWait.cpp
index fada532..8a2cc44 100644
--- a/PxShared/src/cudamanager/src/BlockingWait.cpp
+++ b/PxShared/src/cudamanager/src/BlockingWait.cpp
@@ -63,8 +63,9 @@ void PxGpuWorkerThread::blockingWaitFunc()
 		}
 		else if (!mFailureDetected)
 		{
-			emitStartEvent("GpuDispatcher.BlockingWaitEvent");
-
+#if PX_SUPPORT_PXTASK_PROFILING
+			PX_PROFILE_ZONE("GpuDispatcher.BlockingWaitEvent", 0);
+#endif
 			if (1 & ~intptr_t(b.blockingStream))
 			{
 				GD_CHECK_CALL(cuStreamSynchronize(b.blockingStream));
@@ -73,8 +74,6 @@ void PxGpuWorkerThread::blockingWaitFunc()
 			{
 				GD_CHECK_CALL(cuEventSynchronize(b.blockingEvent));
 			}
-
-			emitStopEvent("GpuDispatcher.BlockingWaitEvent");
 		}
 
 		if (b.blockingEvent)
diff --git a/PxShared/src/cudamanager/src/CudaContextManager.cpp b/PxShared/src/cudamanager/src/CudaContextManager.cpp
index b5b6efc..e05911e 100644
--- a/PxShared/src/cudamanager/src/CudaContextManager.cpp
+++ b/PxShared/src/cudamanager/src/CudaContextManager.cpp
@@ -23,7 +23,7 @@
 // components in life support devices or systems without express written approval of
 // NVIDIA Corporation.
 //
-// Copyright (c) 2008-2014 NVIDIA Corporation. All rights reserved.
+// Copyright (c) 2008-2017 NVIDIA Corporation. All rights reserved.
 
 #include "foundation/PxAssert.h"
 #include "foundation/PxErrorCallback.h"
@@ -88,102 +88,6 @@ static void* GetProcAddress(void* handle, const char* name) { return dlsym(handl
 
 #define ENABLE_DEVICE_INFO_BRINGUP	0
 
-#include "GPUProfile.h"
-
-#if ENABLE_CUDA_DEVICE_RESET
-#include "cudaProfiler.h"
-#endif
-
-#if USE_PERFKIT
-#pragma warning (push)
-#pragma warning (disable : 4099)
-#pragma warning (disable : 4191)
-#define NVPM_INITGUID
-#include <stdio.h>
-#include "cuda.h"
-#include "../../../../../../../externals/nvPerfKit/4.1.0.14260/inc/NvPmApi.Manager.h"
-static NvPmApiManager S_NVPMManager;
-extern NvPmApiManager *GetNvPmApiManager() {return &S_NVPMManager;}
-const NvPmApi *GetNvPmApi() {return S_NVPMManager.Api();}
-NVPMContext hNVPMContext(0);
-
-void initPerfKit()
-{
-	//Sync with GPU
-	cuCtxSynchronize();
-	
-	// Reset counters
-	uint32_t nCount;
-	GetNvPmApi()->Sample(hNVPMContext, NULL, &nCount);
-}
-
-void endPerfKit()
-{
-	//Sync with GPU
-	cuCtxSynchronize();
-	
-	uint32_t nCount;
-	GetNvPmApi()->Sample(hNVPMContext, NULL, &nCount);
-	
-	uint64_t value;
-	uint64_t cycle;
-	
-	uint64_t sum = 0;
-	uint64_t maxVal = 0;
-	char name[512];
-
-	int nvStatus = 0;
-
-	PX_UNUSED(value);
-	PX_UNUSED(cycle);
-	PX_UNUSED(sum);
-	PX_UNUSED(maxVal);
-	PX_UNUSED(name);
-	PX_UNUSED(nvStatus);
-
-	printf("counters:\n");
-
-#if COUNT_L2_TO_L1_BYTES
-	nvStatus |= GetNvPmApi()->GetCounterValueByName(hNVPMContext, "l2_read_bytes", 0, &value, &cycle);
-	printf("L2->L1 bytes %d\n",value);
-#elif COUNT_SM_TO_L1_QUERIES
-	nvStatus |= GetNvPmApi()->GetCounterValueByName(hNVPMContext, "tex_cache_sector_queries", 0, &value, &cycle);
-	printf("SM->L1 queries %d\n",value);
-#endif
-
-#if COUNT_INST_EXECUTED || COUNT_STORE_INST_EXECUTED || COUNT_ACTIVE_CYCLES || COUNT_ACTIVE_WARPS
-	for (int i = 0; i != SM_COUNT; i++)
-	{
-#if COUNT_INST_EXECUTED
-		sprintf_s(name, 512, "sm_inst_executed_vsm%d", i);
-#elif COUNT_STORE_INST_EXECUTED
-		sprintf_s(name, 512, "sm_inst_executed_global_stores_vsm%d", i);
-#elif COUNT_ACTIVE_CYCLES
-		sprintf_s(name, 512, "sm_active_cycles_vsm%d", i);
-#elif COUNT_ACTIVE_WARPS
-		sprintf_s(name, 512, "sm_active_warps_vsm%d", i);
-#endif
-		nvStatus |= GetNvPmApi()->GetCounterValueByName(hNVPMContext, name, 0, &value, &cycle);
-		
-		sum += value;
-		maxVal = physx::PxMax(maxVal, value);
-	}
-#if COUNT_ACTIVE_CYCLES
-	printf("sum %I64d\n", sum);
-#else
-	printf("sum %I64d\n", sum);
-#endif
-
-	if (!nvStatus)
-	{
-		PX_ASSERT(false);
-	}
-#endif
-}
-
-#pragma warning (pop)
-#endif
-
 namespace physx
 {
 
@@ -201,7 +105,7 @@ public:
 	~CudaCtxMgr();
 
 	bool            safeDelayImport(PxErrorCallback& errorCallback);
-	CUcontext       acquireContext();
+	void            acquireContext();
 	void            releaseContext();
 
 	/* All these methods can be called without acquiring the context */
@@ -219,6 +123,7 @@ public:
 	bool            supportsArchSM35() const;  // GK110
 	bool            supportsArchSM50() const;  // GM100
 	bool            supportsArchSM52() const;  // GM200
+	bool            supportsArchSM60() const;  // GP100
 	bool            isIntegrated() const;      // true if GPU is integrated (MCP) part
 	bool            canMapHostMemory() const;  // true if GPU map host memory to GPU
 	int             getDriverVersion() const;
@@ -230,6 +135,7 @@ public:
 	unsigned int	getClockRate() const;
 
 	const char*     getDeviceName() const;
+	CUdevice		getDevice() const;
 	const CUdevprop* getDeviceProperties() const;
 
 	PxCudaInteropMode::Enum	getInteropMode() const;
@@ -251,6 +157,8 @@ public:
 
 	void            release();
 
+	CUcontext		getContext() { return mCtx; }
+
 private:
 
 	int             mSceneCount;
@@ -330,6 +238,10 @@ bool CudaCtxMgr::supportsArchSM52() const
 {
 	return mIsValid && ((mComputeCapMajor > 5) || (mComputeCapMajor == 5 && mComputeCapMinor >= 2));
 }
+bool CudaCtxMgr::supportsArchSM60() const
+{
+	return mIsValid && mComputeCapMajor >= 6;
+}
 
 bool CudaCtxMgr::isIntegrated() const
 {
@@ -380,6 +292,18 @@ const char* CudaCtxMgr::getDeviceName() const
 	}
 }
 
+CUdevice CudaCtxMgr::getDevice() const
+{
+	if (mIsValid)
+	{
+		return mDevHandle;
+	}
+	else
+	{
+		return -1;
+	}
+}
+
 PxCudaInteropMode::Enum CudaCtxMgr::getInteropMode() const
 {
 	return mInteropMode;
@@ -504,8 +428,7 @@ CudaCtxMgr::CudaCtxMgr(const PxCudaContextManagerDesc& desc, PxErrorCallback& er
 
 		if (PhysXDeviceSettings::isUsingDedicatedGPU() == 1 || sliEnabled)
 		{
-			if (mInteropMode == PxCudaInteropMode::D3D9_INTEROP ||
-				mInteropMode == PxCudaInteropMode::D3D10_INTEROP ||
+			if (mInteropMode == PxCudaInteropMode::D3D10_INTEROP ||
 				mInteropMode == PxCudaInteropMode::D3D11_INTEROP)
 			{
 				mInteropMode = PxCudaInteropMode::NO_INTEROP;
@@ -562,18 +485,6 @@ CudaCtxMgr::CudaCtxMgr(const PxCudaContextManagerDesc& desc, PxErrorCallback& er
 			mOwnContext = true;
 		}
 #if PX_WIN32 || PX_WIN64
-		else if (mInteropMode == PxCudaInteropMode::D3D9_INTEROP)
-		{
-			status = cuD3D9CtxCreate(&mCtx, &mDevHandle, (unsigned int)flags,
-			                         reinterpret_cast<IDirect3DDevice9*>(desc.graphicsDevice));
-
-			if (CUDA_SUCCESS != status)
-			{
-				errorCallback.reportError(PxErrorCode::eDEBUG_WARNING, "cuD3D9CtxCreate failed",__FILE__,__LINE__);
-				return;
-			}
-			mOwnContext = true;
-		}
 		else if (mInteropMode == PxCudaInteropMode::D3D10_INTEROP)
 		{
 			status = cuD3D10CtxCreate(&mCtx, &mDevHandle, (unsigned int)flags,
@@ -710,73 +621,6 @@ CudaCtxMgr::CudaCtxMgr(const PxCudaContextManagerDesc& desc, PxErrorCallback& er
 	errorCallback.reportError(PxErrorCode::eDEBUG_INFO, "Number of SM: %d", mMultiprocessorCount);
 	errorCallback.reportError(PxErrorCode::eDEBUG_INFO, "Max Threads Per Block: %d", mMaxThreadsPerBlock);
 #endif
-
-#if USE_PERFKIT
-	{
-#if _WIN64
-		wchar_t * dllName = L"..\\..\\..\\..\\..\\externals\\nvPerfKit\\4.1.0.14260\\bin\\win7_x64\\NvPmApi.Core.dll";
-#else
-		wchar_t * dllName = L"..\\..\\..\\..\\..\\externals\\nvPerfKit\\4.1.0.14260\\bin\\win7_x86\\NvPmApi.Core.dll";
-#endif
-
-		NVPMRESULT nvResult;
-
-		if ((nvResult = GetNvPmApiManager()->Construct(dllName)) != NVPM_OK)
-		{
-			printf("perfkit error 1\n");
-			return; 
-		}
-
-		if ((nvResult = GetNvPmApi()->Init()) != NVPM_OK)
-		{
-			printf("perfkit error 2\n");
-			return; 
-		}
-
-		acquireContext();
-
-		CUcontext ctx;
-		cuCtxGetCurrent(&ctx);
-		if ((nvResult = GetNvPmApi()->CreateContextFromCudaContext((APIContextHandle)ctx, &hNVPMContext)) != NVPM_OK)
-		{
-			printf("perfkit error 3\n");
-			return; // This is an error condition
-		}
-
-		uint32_t nvStatus = 0;
-
-#if COUNT_L2_TO_L1_BYTES
-		nvStatus |= GetNvPmApi()->AddCounterByName(hNVPMContext, "l2_read_bytes");
-#elif COUNT_SM_TO_L1_QUERIES
-		nvStatus |= GetNvPmApi()->AddCounterByName(hNVPMContext, "tex_cache_sector_queries");
-#endif
-
-#if COUNT_INST_EXECUTED || COUNT_STORE_INST_EXECUTED || COUNT_ACTIVE_CYCLES || COUNT_ACTIVE_WARPS
-		char name[512];
-		for (int i = 0; i != SM_COUNT; i++)
-		{
-#if COUNT_INST_EXECUTED
-			sprintf_s(name,512,"sm_inst_executed_vsm%d",i);
-#elif COUNT_STORE_INST_EXECUTED
-			sprintf_s(name, 512, "sm_inst_executed_global_stores_vsm%d",i);
-#elif COUNT_ACTIVE_CYCLES
-			sprintf_s(name, 512, "sm_active_cycles_vsm%d",i);
-#elif COUNT_ACTIVE_WARPS
-			sprintf_s(name, 512, "sm_active_warps_vsm%d",i);
-#endif
-			nvStatus |= GetNvPmApi()->AddCounterByName(hNVPMContext, name);
-		}
-#elif COUNT_GPU_BUSY
-		nvStatus |= GetNvPmApi()->AddCounterByName(hNVPMContext, "gpu_busy");
-#endif
-
-		if (nvStatus != 0)
-		{
-			printf("perfkit error 4\n");
-			return; // This is an error condition
-		}
-	}
-#endif
 }
 
 /* Some driver version mismatches can cause delay import crashes.  Load NVCUDA.dll
@@ -871,10 +715,6 @@ CudaCtxMgr::~CudaCtxMgr()
 	if(!--mManagerRefCount)
 		 shdfnd::TlsFree(mContextRefCountTls);
 #endif
-
-#if ENABLE_CUDA_DEVICE_RESET
-	CUT_SAFE_CALL(cuProfilerStop());
-#endif
 }
 
 bool CudaCtxMgr::registerResourceInCudaGL(CUgraphicsResource& resource, uint32_t buffer, PxCudaInteropRegisterFlags flags)
@@ -900,9 +740,6 @@ bool CudaCtxMgr::registerResourceInCudaD3D(CUgraphicsResource& resource, void* r
 
 	switch (mInteropMode)
 	{
-	case PxCudaInteropMode::D3D9_INTEROP:
-		ret = cuGraphicsD3D9RegisterResource(&resource, (IDirect3DResource9*)resourcePointer, uint32_t(flags));
-		break;
 	case PxCudaInteropMode::D3D10_INTEROP:
 		ret = cuGraphicsD3D10RegisterResource(&resource, (ID3D10Resource*)resourcePointer, uint32_t(flags));
 		break;
@@ -938,7 +775,7 @@ bool CudaCtxMgr::unregisterResourceInCuda(CUgraphicsResource resource)
 	return ret == CUDA_SUCCESS;
 }
 
-CUcontext CudaCtxMgr::acquireContext()
+void CudaCtxMgr::acquireContext()
 {
 	CUcontext ctx = 0;
 	CUT_SAFE_CALL(cuCtxGetCurrent(&ctx));
@@ -955,8 +792,6 @@ CUcontext CudaCtxMgr::acquireContext()
 	char* refCount = (char*)shdfnd::TlsGet(mContextRefCountTls);
 	shdfnd::TlsSet(mContextRefCountTls, ++refCount);
 #endif
-
-	return mCtx;
 }
 
 void CudaCtxMgr::releaseContext()
diff --git a/PxShared/src/cudamanager/src/GpuDispatcher.cpp b/PxShared/src/cudamanager/src/GpuDispatcher.cpp
index 0d05a97..432a0cd 100644
--- a/PxShared/src/cudamanager/src/GpuDispatcher.cpp
+++ b/PxShared/src/cudamanager/src/GpuDispatcher.cpp
@@ -487,21 +487,6 @@ PxGpuWorkerThread::~PxGpuWorkerThread()
 	}
 }
 
-void PxGpuWorkerThread::emitStartEvent(const char *id)
-{
-	PX_UNUSED(id);
-#if PX_SUPPORT_PXTASK_PROFILING
-	PX_PROFILE_START_CROSSTHREAD(id,0);
-#endif
-}
-
-void PxGpuWorkerThread::emitStopEvent(const char *id)
-{
-	PX_UNUSED(id);
-#if PX_SUPPORT_PXTASK_PROFILING
-	PX_PROFILE_STOP_CROSSTHREAD(id,0);
-#endif
-}
 
 /* A TaskManager is informing us that its simulation is being stepped */
 void PxGpuWorkerThread::startSimulation()
@@ -555,15 +540,14 @@ void PxGpuWorkerThread::execute()
  */
 void PxGpuWorkerThread::addCompletionPrereq(PxBaseTask& task)
 {
-	if (mFailureDetected)
-	{
+	if(mFailureDetected)
 		return;
-	}
 
-	emitStartEvent("GpuDispatcher.AddCompletionEvent");
+#if PX_SUPPORT_PXTASK_PROFILING
+	PX_PROFILE_ZONE("GpuDispatcher.AddCompletionEvent", task.getContextId());
+#endif
 	task.addReference();
 	mCompletionTasks.pushBack(&task);
-	emitStopEvent("GpuDispatcher.AddCompletionEvent");
 }
 
 namespace
@@ -757,7 +741,9 @@ void PxGpuWorkerThread::pollSubmitted(shdfnd::Array<ReadyTask>* ready)
 
 void PxGpuWorkerThread::processActiveTasks()
 {
-	emitStartEvent("GpuDispatcher.ProcessTasksEvent");
+#if PX_SUPPORT_PXTASK_PROFILING
+	PX_PROFILE_ZONE("GpuDispatcher.ProcessTasksEvent", 0);	// PT: TODO: fix invalid context
+#endif
 	
 	if (mFailureDetected)
 	{
@@ -766,7 +752,6 @@ void PxGpuWorkerThread::processActiveTasks()
 			mInputReady.reset();
 			mSubmittedTaskList.popBack()->release();
 		}
-		emitStopEvent("GpuDispatcher.ProcessTasksEvent");
 		return;
 	}
 
@@ -824,13 +809,16 @@ void PxGpuWorkerThread::processActiveTasks()
 				else
 				{
 					const CUstream s = (r.task->mStreamIndex > 0) ? mCachedStreams.get(r.task->mStreamIndex) : 0;
+
+					bool active;
+					{
 #if PX_PROFILE
-					r.task->mTm->emitStartEvent(*r.task);
+#if PX_SUPPORT_PXTASK_PROFILING
+						PX_PROFILE_ZONE(r.task->getName(), r.task->getContextId());
 #endif
-					bool active = r.task->launchInstance(s, int(r.iteration++));
-#if PX_PROFILE
-					r.task->mTm->emitStopEvent(*r.task);
 #endif
+						active = r.task->launchInstance(s, int(r.iteration++));
+					}
 					if(singleStream != r.task->mStreamIndex)
 						singleStream = 0;
 
@@ -935,8 +923,6 @@ void PxGpuWorkerThread::processActiveTasks()
 	while (tasksRemain);
 
 	mCachedNonBlockingEvents.add(nonBlockEv);
-
-	emitStopEvent("GpuDispatcher.ProcessTasksEvent");
 }
 
 #endif
diff --git a/PxShared/src/foundation/include/PsAllocator.h b/PxShared/src/foundation/include/PsAllocator.h
index cbf32d3..f988b6c 100644
--- a/PxShared/src/foundation/include/PsAllocator.h
+++ b/PxShared/src/foundation/include/PsAllocator.h
@@ -108,7 +108,7 @@
 #elif PX_XBOXONE
 #include <malloc.h>
 #define PxAlloca(x) alloca(x)
-#elif PX_NX
+#elif PX_SWITCH
 #include <malloc.h>
 #define PxAlloca(x) alloca(x)
 #endif
diff --git a/PxShared/src/foundation/include/PsAoS.h b/PxShared/src/foundation/include/PsAoS.h
index 641a40a..5a7c82d 100644
--- a/PxShared/src/foundation/include/PsAoS.h
+++ b/PxShared/src/foundation/include/PsAoS.h
@@ -34,7 +34,7 @@
 
 #if PX_WINDOWS && !PX_NEON
 #include "windows/PsWindowsAoS.h"
-#elif(PX_UNIX_FAMILY || PX_PS4 || PX_NX)
+#elif(PX_UNIX_FAMILY || PX_PS4 || PX_SWITCH)
 #include "unix/PsUnixAoS.h"
 #elif PX_XBOXONE
 #include "XboxOne/PsXboxOneAoS.h"
diff --git a/PxShared/src/foundation/include/PsInlineAoS.h b/PxShared/src/foundation/include/PsInlineAoS.h
index 6d43607..6ae15cf 100644
--- a/PxShared/src/foundation/include/PsInlineAoS.h
+++ b/PxShared/src/foundation/include/PsInlineAoS.h
@@ -35,7 +35,7 @@
 #if PX_WINDOWS
 #include "windows/PsWindowsTrigConstants.h"
 #include "windows/PsWindowsInlineAoS.h"
-#elif(PX_UNIX_FAMILY || PX_PS4 || PX_NX)
+#elif(PX_UNIX_FAMILY || PX_PS4 || PX_SWITCH)
 #include "unix/PsUnixTrigConstants.h"
 #include "unix/PsUnixInlineAoS.h"
 #elif PX_XBOXONE
diff --git a/PxShared/src/foundation/include/PsIntrinsics.h b/PxShared/src/foundation/include/PsIntrinsics.h
index 1e1b9d1..38b91ba 100644
--- a/PxShared/src/foundation/include/PsIntrinsics.h
+++ b/PxShared/src/foundation/include/PsIntrinsics.h
@@ -38,8 +38,8 @@
 #include "unix/PsUnixIntrinsics.h"
 #elif PX_XBOXONE
 #include "XboxOne/PsXboxOneIntrinsics.h"
-#elif PX_NX
-#include "nx/PsNXIntrinsics.h"
+#elif PX_SWITCH
+#include "switch/PsSwitchIntrinsics.h"
 #else
 #error "Platform not supported!"
 #endif
diff --git a/PxShared/src/foundation/include/PsThread.h b/PxShared/src/foundation/include/PsThread.h
index 8ba553a..4e7c104 100644
--- a/PxShared/src/foundation/include/PsThread.h
+++ b/PxShared/src/foundation/include/PsThread.h
@@ -41,7 +41,7 @@
 
 #if PX_WINDOWS_FAMILY || PX_XBOXONE
 #define PxSpinLockPause() __asm pause
-#elif PX_LINUX || PX_ANDROID || PX_PS4 || PX_APPLE_FAMILY || PX_NX
+#elif PX_LINUX || PX_ANDROID || PX_PS4 || PX_APPLE_FAMILY || PX_SWITCH
 #define PxSpinLockPause() asm("nop")
 #else
 #error "Platform not supported!"
@@ -173,7 +173,7 @@ class PX_FOUNDATION_API ThreadImpl
 	Change the affinity mask for this thread. The mask is a platform
 	specific value.
 
-	On Windows, Linux, PS4, XboxOne and NX platforms, each set mask bit represents
+	On Windows, Linux, PS4, XboxOne and Switch platforms, each set mask bit represents
 	the index of a logical processor that the OS may schedule thread execution on.
 	Bits outside the range of valid logical processors may be ignored or cause
 	the function to return an error.
diff --git a/PxShared/src/foundation/include/PsVecMath.h b/PxShared/src/foundation/include/PsVecMath.h
index 4e891d8..ffd2de8 100644
--- a/PxShared/src/foundation/include/PsVecMath.h
+++ b/PxShared/src/foundation/include/PsVecMath.h
@@ -54,7 +54,7 @@
 #define COMPILE_VECTOR_INTRINSICS 1
 #elif PX_IOS&& PX_NEON
 #define COMPILE_VECTOR_INTRINSICS 1
-#elif PX_NX
+#elif PX_SWITCH
 #define COMPILE_VECTOR_INTRINSICS 1
 #else
 #define COMPILE_VECTOR_INTRINSICS 0
diff --git a/PxShared/src/foundation/include/PsVecMathAoSScalarInline.h b/PxShared/src/foundation/include/PsVecMathAoSScalarInline.h
index 9bef465..f5dea7b 100644
--- a/PxShared/src/foundation/include/PsVecMathAoSScalarInline.h
+++ b/PxShared/src/foundation/include/PsVecMathAoSScalarInline.h
@@ -40,8 +40,13 @@
 
 #define BOOL_TO_U16(b) (PxU16)(- PxI32(b))
 
+#define PX_VECMATH_ASSERT_ENABLED 0
 
+#if PX_VECMATH_ASSERT_ENABLED
 #define VECMATHAOS_ASSERT(x) { PX_ASSERT(x); }
+#else
+#define VECMATHAOS_ASSERT(x)
+#endif
 
 /////////////////////////////////////////////////////////////////////
 ////INTERNAL USE ONLY AND TESTS
@@ -1479,7 +1484,7 @@ PX_FORCE_INLINE BoolV BOr(const BoolV a, const BoolV b)
 
 PX_FORCE_INLINE PxU32 BAllEq(const BoolV a, const BoolV b)
 {
-	return (a.ux == b.ux && a.uy == b.uy && a.uz == b.uz && a.uw == b.uw ? TRUE_TO_U32 : FALSE_TO_U32);
+	return (a.ux == b.ux && a.uy == b.uy && a.uz == b.uz && a.uw == b.uw ? 1 : 0);
 }
 
 PX_FORCE_INLINE PxU32 BAllEqTTTT(const BoolV a)
diff --git a/PxShared/src/foundation/include/unix/neon/PsUnixNeonInlineAoS.h b/PxShared/src/foundation/include/unix/neon/PsUnixNeonInlineAoS.h
index 2a0578d..a97f821 100644
--- a/PxShared/src/foundation/include/unix/neon/PsUnixNeonInlineAoS.h
+++ b/PxShared/src/foundation/include/unix/neon/PsUnixNeonInlineAoS.h
@@ -42,7 +42,7 @@
 
 // "exact"
 #define VRECIPQ recipq_newton<4>
-#if PX_NX
+#if PX_SWITCH
 // StabilizationTests.AveragePoint needs more precision to succeed.
 #define VRECIP recip_newton<5> 
 #else
diff --git a/PxShared/src/foundation/src/PsAssert.cpp b/PxShared/src/foundation/src/PsAssert.cpp
index 3070383..295a81e 100644
--- a/PxShared/src/foundation/src/PsAssert.cpp
+++ b/PxShared/src/foundation/src/PsAssert.cpp
@@ -34,8 +34,8 @@
 
 #if PX_WINDOWS_FAMILY
 #include <crtdbg.h>
-#elif PX_NX
-#include "nx/PsNXAbort.h"
+#elif PX_SWITCH
+#include "switch/PsSwitchAbort.h"
 #endif
 
 namespace
@@ -63,7 +63,7 @@ class DefaultAssertHandler : public physx::PxAssertHandler
 		__debugbreak();
 #elif PX_WINDOWS_FAMILY&& PX_CHECKED
 		__debugbreak();
-#elif PX_NX
+#elif PX_SWITCH
 		abort(buffer);
 #else
 		abort();
diff --git a/PxShared/src/pvd/include/PxProfileAllocatorWrapper.h b/PxShared/src/pvd/include/PxProfileAllocatorWrapper.h
index 22903ec..abac561 100644
--- a/PxShared/src/pvd/include/PxProfileAllocatorWrapper.h
+++ b/PxShared/src/pvd/include/PxProfileAllocatorWrapper.h
@@ -74,7 +74,7 @@ namespace physx { namespace profile {
 	{
 		static const char* getName()
 		{
-#if PX_LINUX || PX_ANDROID || PX_PS4 || PX_IOS || PX_OSX || PX_EMSCRIPTEN || PX_NX
+#if PX_LINUX || PX_ANDROID || PX_PS4 || PX_IOS || PX_OSX || PX_EMSCRIPTEN || PX_SWITCH
 			return __PRETTY_FUNCTION__;
 #else
 			return typeid(T).name();
diff --git a/PxShared/src/task/src/TaskManager.cpp b/PxShared/src/task/src/TaskManager.cpp
index ffcbfcd..c6210bd 100644
--- a/PxShared/src/task/src/TaskManager.cpp
+++ b/PxShared/src/task/src/TaskManager.cpp
@@ -168,9 +168,6 @@ public:
 	void	decrReference( PxLightCpuTask& lighttask );
 	void	addReference( PxLightCpuTask& lighttask );		
 
-	void	emitStartEvent( PxBaseTask& basetask, uint32_t threadId=0);
-	void	emitStopEvent( PxBaseTask& basetask, uint32_t threadId=0);
-
 	PxErrorCallback&			mErrorCallback;
 	PxCpuDispatcher           *mCpuDispatcher;
 	PxGpuDispatcher           *mGpuDispatcher;		
@@ -304,35 +301,6 @@ void PxTaskMgr::addReference(PxLightCpuTask& lighttask)
 	shdfnd::atomicIncrement(&lighttask.mRefCount);
 }
 
-void PxTaskMgr::emitStartEvent(PxBaseTask& basetask, uint32_t threadId)
-{
-#if DOT_LOG
-	currentTask = &basetask;
-#endif
-
-	PxBaseTask* tmp = &basetask;
-	PX_UNUSED(tmp);
-	PX_UNUSED(threadId);
-
-	/* This does not need a lock! */
-#if PX_SUPPORT_PXTASK_PROFILING
-	//PX_COMPILE_TIME_ASSERT(sizeof(PxProfileEventId::mEventId) == sizeof(PxBaseTask::mEventID));
-	PX_PROFILE_START_CROSSTHREAD(basetask.getName(),0);
-#endif
-}
-
-void PxTaskMgr::emitStopEvent(PxBaseTask& basetask, uint32_t threadId)
-{
-	PxBaseTask* tmp = &basetask;
-	PX_UNUSED(tmp);
-	PX_UNUSED(threadId);
-
-	/* This does not need a lock! */
-#if PX_SUPPORT_PXTASK_PROFILING
-	//PX_COMPILE_TIME_ASSERT(sizeof(PxProfileEventId::mEventId) == sizeof(PxBaseTask::mEventID));
-	PX_PROFILE_STOP_CROSSTHREAD(basetask.getName(),0);
-#endif
-}
 
 /*
  * Called by the owner (Scene) at the start of every frame, before
author	Marijn Tamis <[email protected]>	2017-07-03 11:49:08 +0200
committer	Marijn Tamis <[email protected]>	2017-07-03 11:49:08 +0200
commit	cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381 (patch)
tree	5cc014922d20561d87105d279b6f7eb3e628c6d9
parent	Fix windows line endings in github. (diff)
download	nvcloth-cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381.tar.xz nvcloth-cfa944ded7370fb5f22b1fb894ecf6b9bd3f7381.zip