ZLUDA and LLama.cpp not loading

### zluda_trace logs (tarball/zip file) _No response_ ### Description get the following error on 13.1: CUDA error: the function requires an architectural feature absent from the device current device: 0, in function ggml_cuda_op_mul_mat_cublas at C:\llamaZLUDA\llama.cpp\ggml\src\ggml-cuda\ggml-cuda.cu:1333 cublasGemmEx(ctx.cublas_handle(id), CUBLAS_OP_T, CUBLAS_OP_N, row_diff, src1_ncols, ne10, &alpha_f16, src0_ptr, CUDA_R_16F, ne00, src1_ptr, CUDA_R_16F, ne10, &beta_f16, dst_f16.get(), CUDA_R_16F, ldc, CUBLAS_COMPUTE_16F, CUBLAS_GEMM_DEFAULT_TENSOR_OP) C:\llamaZLUDA\llama.cpp\ggml\src\ggml-cuda\ggml-cuda.cu:96: CUDA error When i try to use llama.cpp with 12.4 , the whole PC freezes. ### Steps to reproduce currently have 6700xt working with ROCm 7.1.1 Built ZLUDA and LLAMA.cpp : cargo xtask --release cmake -B build -DGGML_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="86" -DGGML_CUDA_FORCE_CUBLAS=true -DLLAMA_CURL=OFF Unable to launch ### ZLUDA version built 15 Jan 26 ### Operating System windows 10 ### GPU 6700xt