/// <summary> /// Invokes the specified kernels. /// </summary> /// <param name="kernels">The kernels.</param> /// <param name="context">The context.</param> /// <param name="cudaContext">The cuda context.</param> /// <param name="result">The result.</param> /// <param name="src">The source.</param> public static void Invoke(FillCopyKernels kernels, TSCudaContext context, CudaContext cudaContext, NDArray result, NDArray src) { var ptx = kernels.GetPtx(context.Compiler); var elementCount = result.ElementCount(); ApplyOpInvoke.Invoke(context, cudaContext, ptx, "copy", result, src, elementCount); }
public static void Invoke(FillCopyKernels kernels, TSCudaContext context, CudaContext cudaContext, Tensor result, Tensor src) { // cudaContext.SetCurrent(); byte[] ptx = kernels.GetPtx(context.Compiler); long elementCount = result.ElementCount(); ApplyOpInvoke.Invoke(context, cudaContext, ptx, "copy", result, src, elementCount); }
/// <summary> /// Invokes the specified kernels. /// </summary> /// <param name="kernels">The kernels.</param> /// <param name="result">The result.</param> /// <param name="value">The value.</param> public static void Invoke(FillCopyKernels kernels, NDArray result, float value) { var context = CudaHelpers.TSContextForTensor(result); var cudaContext = context.CudaContextForTensor(result); var ptx = kernels.GetPtx(context.Compiler); var elementCount = result.ElementCount(); ApplyOpInvoke.Invoke(context, cudaContext, ptx, "fill", result, value, elementCount); }
public static void Invoke(FillCopyKernels kernels, Tensor result, float value) { TSCudaContext context = CudaHelpers.TSContextForTensor(result); CudaContext cudaContext = context.CudaContextForTensor(result); cudaContext.SetCurrent(); byte[] ptx = kernels.GetPtx(context.Compiler); long elementCount = result.ElementCount(); ApplyOpInvoke.Invoke(context, cudaContext, ptx, "fill", result, value, elementCount); }