simpleCooperativeGroups cudaDeviceSynchronize cudaGetErrorString whole ./ ../ ../../../Common Cooperative Groups GPGPU CUDA Parallel Reduction Cooperative Groups true simpleCooperativeGroups.cu 1:CUDA Basic Topics sm50 sm52 sm53 sm60 sm61 sm70 sm72 sm75 sm80 sm86 sm87 sm90 x86_64 linux windows7 x86_64 macosx arm sbsa ppc64le linux all Simple Cooperative Groups exe