This repository has been archived by the owner on Mar 21, 2024. It is now read-only.
thrust::reduce_by_key error for 2^31 elements #1609
Labels
backend: CUDA
Related to the CUDA backend
P1: should have
Necessary, but not critical.
type: bug: functional
Does not work as intended.
Milestone
Tested environment
GA102 (48 GB GPU memory)
CUDA 11.4
Thrust 1.15.0
The code below failed inside the 'reduce_by_key' call with a memory allocation failure; tried to allocate 18446744073694639872 bytes (16 EB) and failed as my system has only 48 GB. This sounds like a bug related to 32 bit signed integer overflow.
The text was updated successfully, but these errors were encountered: