Removed __cuda_callable__ from methods in DistributedNDArray and DistributedNDArrayView
CUDA kernels should not ever work with distributed data structures, they should always get the underlying *local* data structure.
Loading
Please sign in to comment