Skip to content

opt(insert-and-evict): thrust prefix_sum introduce cudaMalloc/cudaFre…#182

Open
Lifann wants to merge 1 commit intoNVIDIA-Merlin:masterfrom
Lifann:opt-insert-and-evict
Open

opt(insert-and-evict): thrust prefix_sum introduce cudaMalloc/cudaFre…#182
Lifann wants to merge 1 commit intoNVIDIA-Merlin:masterfrom
Lifann:opt-insert-and-evict

Conversation

@Lifann
Copy link
Copy Markdown
Collaborator

@Lifann Lifann commented Dec 12, 2023

opt(insert-and-evict): thrust prefix_sum introduces cudaMalloc and cudaFree which make device sync. Replace it by cub API.

The output of unit test case insert-and-evict is as follow:
ut_output.txt

@github-actions
Copy link
Copy Markdown

@Lifann Lifann force-pushed the opt-insert-and-evict branch from 4861ed0 to e728dd2 Compare December 12, 2023 12:07
Comment thread include/merlin_hashtable.cuh Outdated
…e which make host wait. Replace it by cub API.
@Lifann Lifann force-pushed the opt-insert-and-evict branch from e728dd2 to a78f415 Compare December 13, 2023 06:35
@rhdong
Copy link
Copy Markdown
Member

rhdong commented Dec 14, 2023

/blossom-ci

Copy link
Copy Markdown
Member

@rhdong rhdong left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants