pub fn execute_graph<D: KernelDispatch>(
graph: &ComputeGraph,
dispatcher: &mut D,
) -> Result<usize, GpuError>Expand description
Execute a compute graph using the provided kernel dispatcher.
Walks nodes in topological order. Leaf nodes (TensorOp::None) are skipped – they represent input tensors whose data is already on device.
Returns the number of kernel launches performed.