diff --git a/demo/gpudemo/GpuDemo.h b/demo/gpudemo/GpuDemo.h
index d4d7e8d9b..f61f97501 100644
--- a/demo/gpudemo/GpuDemo.h
+++ b/demo/gpudemo/GpuDemo.h
@@ -37,9 +37,9 @@ public:
                     :useOpenCL(true),
                     preferredOpenCLPlatformIndex(-1),
                     preferredOpenCLDeviceIndex(-1),
-					arraySizeX(10),
-		arraySizeY(20),
-		arraySizeZ(10),
+					arraySizeX(40),
+		arraySizeY(40),
+		arraySizeZ(40),
 		m_useConcaveMesh(false),
 		gapX(14.3),
 		gapY(14.0),
diff --git a/demo/gpudemo/main_opengl3core.cpp b/demo/gpudemo/main_opengl3core.cpp
index 4907e32b4..eb757d407 100644
--- a/demo/gpudemo/main_opengl3core.cpp
+++ b/demo/gpudemo/main_opengl3core.cpp
@@ -381,7 +381,8 @@ int main(int argc, char* argv[])
 	
 	args.GetCmdLineArgument("selected_demo",selectedDemo);
 
-	useNewBatchingKernel = args.CheckCmdLineFlag("new_batching");
+	if (args.CheckCmdLineFlag("new_batching"))
+		useNewBatchingKernel = true;
 	bool benchmark=args.CheckCmdLineFlag("benchmark");
 	dump_timings=args.CheckCmdLineFlag("dump_timings");
 	ci.useOpenCL = !args.CheckCmdLineFlag("disable_opencl");
diff --git a/opencl/gpu_rigidbody/host/Solver.cpp b/opencl/gpu_rigidbody/host/Solver.cpp
index b63468160..f9a54fe12 100644
--- a/opencl/gpu_rigidbody/host/Solver.cpp
+++ b/opencl/gpu_rigidbody/host/Solver.cpp
@@ -17,7 +17,7 @@ subject to the following restrictions:
 #include "Solver.h"
 
 ///useNewBatchingKernel  is a rewritten kernel using just a single thread of the warp, for experiments
-bool useNewBatchingKernel = false;
+bool useNewBatchingKernel = true;
 
 #define SOLVER_SETUP_KERNEL_PATH "opencl/gpu_rigidbody/kernels/solverSetup.cl"
 #define SOLVER_SETUP2_KERNEL_PATH "opencl/gpu_rigidbody/kernels/solverSetup2.cl"