Sergey Maydanov

BUG: Compute-follows-data breaks on the inverse operation

2

``` x = dpnp.asarray([1.0, 2.0, 3.0]) y = 1.0/x ``` `y` is numpy array vs. dpnp array Complete reproducer is here https://gist.github.com/samaid/b4510dbcd7285ceed36b42dd96acf30c

Missing dpnp.uint8, dpnp.int8, etc.

https://github.com/samaid/mcpi

DPNP documentation says that dpnp.random.choice() function falls back to numpy. https://intelpython.github.io/dpnp/reference/generated/dpnp.random.choice.html#dpnp.random.choice My DPNP example for Game Of Life requires this function to be run on the device.

Since Numpy version 1.17.0 the Generator can be initialized with a number of different BitGenerators. It exposes many different probability distributions. See NEP 19 for context on the updated random...

As a result my PyCharm environment does not see functions and their signatures in codes like this ` import dpnp as np np.random.seed(111) ` Numpy behaves differently, and the following...

Tried to work with NUMBA-DPPY on Windows. It is unusable because of missing DPNP

Extend nvbench to measure SOL for compute-bound workloads

1

Existing nvbench allows to measure SOL for memory bound workloads by providing ``` state.addGlobalMemoryReads(nbytes) state.addGlobalMemoryWrites(nbytes) ``` It would be useful to extend this concept to provide flops such that we...