This repository has been archived by the owner on Mar 12, 2021. It is now read-only.
This repository has been archived by the owner on Mar 12, 2021. It is now read-only.
polish mapreduce
interface #204
Closed
Description
openedon Nov 16, 2018
Don't let the following call to mapreduce
fallback to CPU
using CuArrays, CUDAnative
CuArrays.allowscalar(false)
x = randn(1000) |> cu
CUDAnative.map(CUDAnative.exp, x)
CUDAnative.reduce(+, x)
# works, but no kernel is evaluated
@device_code_warntype CUDAnative.mapreduce(exp, +, x) # no kernel evaluated
# the following code will cause Julia break down
@device_code_warntype CUDAnative.mapreduce(CUDAnative.exp, +, x)
Issue #180 can be helpful in preventing unexpected fallback, with data transferred silently.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Metadata
Assignees
Labels
No labels