diff --git a/bioem_cuda.cu b/bioem_cuda.cu
index e5ccf87bc6ec27f12d3c5a86cde00e84f59f3084..64145117c41337e51260818f62fe13459684ba05 100644
--- a/bioem_cuda.cu
+++ b/bioem_cuda.cu
@@ -288,7 +288,6 @@ int bioem_cuda::selectCudaDevice()
 	}
 	for (int i = 0;i < count;i++)
 	{
-		printf("CUDA device %d\n", i);
 #if CUDA_VERSION > 3010
 		size_t free, total;
 #else
@@ -301,11 +300,9 @@ int bioem_cuda::selectCudaDevice()
 		cuCtxCreate(&tmpContext, 0, tmpDevice);
 		if(cuMemGetInfo(&free, &total)) exit(1);
 		cuCtxDestroy(tmpContext);
-		if (DebugOutput >= 1) printf("Obtained current memory usage for device %d\n", i);
 		checkCudaErrors(cudaGetDeviceProperties(&deviceProp, i));
-		if (DebugOutput >= 1) printf("Obtained device properties for device %d\n", i);
 
-		if (DebugOutput >= 1) printf("%2d: %s (Rev: %d.%d - Mem Avail %lld / %lld)", i, deviceProp.name, deviceProp.major, deviceProp.minor, (long long int) free, (long long int) deviceProp.totalGlobalMem);
+		if (DebugOutput >= 1) printf("CUDA Device %2d: %s (Rev: %d.%d - Mem Avail %lld / %lld)\n", i, deviceProp.name, deviceProp.major, deviceProp.minor, (long long int) free, (long long int) deviceProp.totalGlobalMem);
 		long long int deviceSpeed = (long long int) deviceProp.multiProcessorCount * (long long int) deviceProp.clockRate * (long long int) deviceProp.warpSize;
 		if (deviceSpeed > bestDeviceSpeed)
 		{
@@ -316,24 +313,24 @@ int bioem_cuda::selectCudaDevice()
 
 	cudaGetDeviceProperties(&deviceProp ,bestDevice); 
 
-	if (DebugOutput >= 1)
+	if (DebugOutput >= 2)
 	{
-		printf("Using CUDA Device %s with Properties:", deviceProp.name);
-		printf("totalGlobalMem = %lld", (unsigned long long int) deviceProp.totalGlobalMem);
-		printf("sharedMemPerBlock = %lld", (unsigned long long int) deviceProp.sharedMemPerBlock);
-		printf("regsPerBlock = %d", deviceProp.regsPerBlock);
-		printf("warpSize = %d", deviceProp.warpSize);
-		printf("memPitch = %lld", (unsigned long long int) deviceProp.memPitch);
-		printf("maxThreadsPerBlock = %d", deviceProp.maxThreadsPerBlock);
-		printf("maxThreadsDim = %d %d %d", deviceProp.maxThreadsDim[0], deviceProp.maxThreadsDim[1], deviceProp.maxThreadsDim[2]);
-		printf("maxGridSize = %d %d %d", deviceProp.maxGridSize[0], deviceProp.maxGridSize[1], deviceProp.maxGridSize[2]);
-		printf("totalConstMem = %lld", (unsigned long long int) deviceProp.totalConstMem);
-		printf("major = %d", deviceProp.major);
-		printf("minor = %d", deviceProp.minor);
-		printf("clockRate = %d", deviceProp.clockRate);
-		printf("memoryClockRate = %d", deviceProp.memoryClockRate);
-		printf("multiProcessorCount = %d", deviceProp.multiProcessorCount);
-		printf("textureAlignment = %lld", (unsigned long long int) deviceProp.textureAlignment);
+		printf("Using CUDA Device %s with Properties:\n", deviceProp.name);
+		printf("totalGlobalMem = %lld\n", (unsigned long long int) deviceProp.totalGlobalMem);
+		printf("sharedMemPerBlock = %lld\n", (unsigned long long int) deviceProp.sharedMemPerBlock);
+		printf("regsPerBlock = %d\n", deviceProp.regsPerBlock);
+		printf("warpSize = %d\n", deviceProp.warpSize);
+		printf("memPitch = %lld\n", (unsigned long long int) deviceProp.memPitch);
+		printf("maxThreadsPerBlock = %d\n", deviceProp.maxThreadsPerBlock);
+		printf("maxThreadsDim = %d %d %d\n", deviceProp.maxThreadsDim[0], deviceProp.maxThreadsDim[1], deviceProp.maxThreadsDim[2]);
+		printf("maxGridSize = %d %d %d\n", deviceProp.maxGridSize[0], deviceProp.maxGridSize[1], deviceProp.maxGridSize[2]);
+		printf("totalConstMem = %lld\n", (unsigned long long int) deviceProp.totalConstMem);
+		printf("major = %d\n", deviceProp.major);
+		printf("minor = %d\n", deviceProp.minor);
+		printf("clockRate = %d\n", deviceProp.clockRate);
+		printf("memoryClockRate = %d\n", deviceProp.memoryClockRate);
+		printf("multiProcessorCount = %d\n", deviceProp.multiProcessorCount);
+		printf("textureAlignment = %lld\n", (unsigned long long int) deviceProp.textureAlignment);
 	}
 	
 	if (DebugOutput >= 1)