Skip to content

Remove unimplemented cagra search apis - #2600

Merged
rapids-bot[bot] merged 14 commits into
NVIDIA:mainfrom
aamijar:cagra-vpq-f32-remove
Sep 30, 2026
Merged

rapids-bot[bot] merged 14 commits into
NVIDIA:mainfrom
aamijar:cagra-vpq-f32-remove

Conversation

@aamijar

@aamijar aamijar commented Sep 12, 2026 •

Copy link
Copy Markdown
Member

The cagra search APIs with vpq f32 are not implemented and result in a runtime error. Same goes for cagra search and MG cagra search on standard datasets.
Instead users should get a compile/link time error. We should simply remove the apis and the vpq_f32_index type which doesn't have any use cases currently.

Update: We will keep the cagra search and MG cagra search on standard dataset, and simply implement the path where if the alignment meets cagra's requirements we can call the search function. However, if it doesn't we can give an error.

@aamijar
aamijar requested a review from a team as a code owner September 12, 2026 18:54
@aamijar aamijar self-assigned this Sep 12, 2026
@aamijar aamijar added non-breaking Introduces a non-breaking change improvement Improves an existing functionality labels Sep 12, 2026
@aamijar aamijar moved this to In Progress in Unstructured Data Processing Sep 12, 2026
@aamijar
aamijar requested a review from a team as a code owner September 12, 2026 19:29
@aamijar aamijar changed the title Remove unimplemented cagra search vpq f32 apis Remove unimplemented cagra search apis Sep 12, 2026
raft::device_matrix_view<float, int64_t, raft::row_major> distances,
const cuvs::neighbors::filtering::base_filter& sample_filter =
cuvs::neighbors::filtering::none_sample_filter{});
void search(raft::resources const& res,

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

we can leave the device_standard_index APIs as is. We can create the padded view ourselves if needed. If the dim is not aligned, only then we can error out.

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

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

Alright I'll change it so that we call the search kernel if its already aligned, instead of always erroring out.

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

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

Addressed in 9d02b04

@aamijar aamijar moved this from In Progress to Needs Review in Unstructured Data Processing Sep 24, 2026
@aamijar

aamijar commented Sep 30, 2026

Copy link
Copy Markdown
Member Author

/merge

@rapids-bot
rapids-bot Bot merged commit d3df668 into NVIDIA:main Sep 30, 2026
197 of 199 checks passed
maxwbuckley added a commit to maxwbuckley/cuvs that referenced this pull request Oct 6, 2026
Upstream moved cagra::search_params earlier in cagra.hpp (NVIDIA#2600). Resolved by
keeping main's layout and dropping the duplicate filtering_rate field from it,
since it is now inherited from cuvs::neighbors::search_params; upstream's UDF
fallback documentation is carried into the replacement comment.

The base-struct doc now says the popcount + sync applies to bitset_filter,
since CAGRA's bloom and roaring filters derive the rate without one.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

improvement Improves an existing functionality non-breaking Introduces a non-breaking change

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

2 participants