k-Wave User Forum » Topic: Compiling the code with cuda8 on laptop

k-Wave User Forum » Topic: Compiling the code with cuda8 on laptop - GeForce GTX 960M http://www.k-wave.org/forum/topic/compiling-the-code-with-cuda8-on-laptop-geforce-gtx-960m Support for the k-Wave MATLAB toolbox en-US Tue, 12 May 2026 23:35:22 +0000 http://bbpress.org/?v=1.0.2 <![CDATA[Search]]> q http://www.k-wave.org/forum/search.php luca.forte on "Compiling the code with cuda8 on laptop - GeForce GTX 960M" http://www.k-wave.org/forum/topic/compiling-the-code-with-cuda8-on-laptop-geforce-gtx-960m#post-5874 Thu, 30 Mar 2017 20:59:43 +0000 luca.forte 5874@http://www.k-wave.org/forum/ Hello Jiri, thank you so much. You are all very kind and professional. I'll wait until a new version is available and I'll get back to you. Thanks, Luca. Jiri Jaros on "Compiling the code with cuda8 on laptop - GeForce GTX 960M" http://www.k-wave.org/forum/topic/compiling-the-code-with-cuda8-on-laptop-geforce-gtx-960m#post-5871 Thu, 30 Mar 2017 11:30:36 +0000 Jiri Jaros 5871@http://www.k-wave.org/forum/ Hi luca.forte, I'm working on it. CUDA 8.0 changed the data layout for FFT. It's going to take some time me to fix all kernels. There are two tweaks you can use now to get a limited functionality. 1. Compile it with CUDA 7.5 and remove these lines form the makefile --generate-code arch=compute_20,code=sm_20 \ --generate-code arch=compute_20,code=sm_21 \ --generate-code arch=compute_30,code=sm_30 \ --generate-code arch=compute_32,code=sm_32 \ --generate-code arch=compute_35,code=sm_35 \ --generate-code arch=compute_37,code=sm_37 \ --generate-code arch=compute_50,code=sm_50 \ --generate-code arch=compute_52,code=sm_52 \ --generate-code arch=compute_53,code=sm_53 \ 2. If you don't need --u_non_staggered_raw, just comment out these lines cufftError = cufftSetCompatibilityMode(cufftPlan_1DY_C2R, CUFFT_COMPATIBILITY_NATIVE); 3. If you can wait till the end of April, I will release a fixed version. Jiri luca.forte on "Compiling the code with cuda8 on laptop - GeForce GTX 960M" http://www.k-wave.org/forum/topic/compiling-the-code-with-cuda8-on-laptop-geforce-gtx-960m#post-5870 Mon, 27 Mar 2017 04:07:21 +0000 luca.forte 5870@http://www.k-wave.org/forum/ Hello, I'm trying to compile the 3Dcode, but I get the following message: MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_3D_R2C(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:113:60: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_3D_R2C, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_3D_C2R(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:139:61: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_3D_C2R, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DX_R2C(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:189:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DX_R2C, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DY_R2C(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:238:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DY_R2C, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DZ_R2C(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:285:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DZ_R2C, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DX_C2R(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:335:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DX_C2R, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DY_C2R(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:383:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DY_C2R, CUFFT_COMPATIBILITY_NATIVE); ^ MatrixClasses/CUFFTComplexMatrix.cpp: In static member function ‘static void TCUFFTComplexMatrix::Create_FFT_Plan_1DZ_C2R(const TDimensionSizes&)’: MatrixClasses/CUFFTComplexMatrix.cpp:432:62: error: ‘CUFFT_COMPATIBILITY_NATIVE’ was not declared in this scope cufftError = cufftSetCompatibilityMode(cufftPlan_1DZ_C2R, CUFFT_COMPATIBILITY_NATIVE); ^ <builtin>: recipe for target 'MatrixClasses/CUFFTComplexMatrix.o' failed make: *** [MatrixClasses/CUFFTComplexMatrix.o] Error 1 Can anyone help me? Thanks, Luca.