NVIDIA / NVIDIA/cccl

[Documentation] Behavior of block scans with initial value param

Open
#859 5 comments 0 reactions 1 assignee Claimed by @gonidelis View on GitHub
cub
Dominant language
C++
Stars
2.5k
Forks
486
Avg merge
2d 6h
Merged PRs (30d)
295

Description

Hi,
I spent quite some time chasing down a bug which ended up being a misunderstanding of the behavior of block scans when an initial val param is given. It turns out that the block aggregate does not include the initial val. i.e., The exclusive sum of {1,2,3} with initial val 10 is {10, 11, 13 }, but the block aggregate is not 16. It is 6. This was not clear from the documentation and the general expectation of an aggregate.

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.