Peter Andreas EntschevinRAPIDS AIHigh-Performance Python Communication with UCX-PyUCX-Py is an accelerated networking library designed for low-latency high bandwidth transfers for both host and GPU device memory objects.Apr 15, 2020Apr 15, 2020
Peter Andreas EntschevinRAPIDS AIParallelizing Custom CuPy Kernels with DaskUser-defined, custom kernels in Chainer CuPyAug 15, 2019Aug 15, 2019
Peter Andreas EntschevinRAPIDS AISingle-GPU CuPy SpeedupsSee Chainer’s CuPy performance for various data science operations running on a single GPUJul 23, 2019Jul 23, 2019