[REVIEW] New Dataset API Clarifying Ownership by HowardHuang1 · Pull Request #1846 · NVIDIA/cuvs

HowardHuang1 · 2026-02-24T23:21:21Z

Overview

Addressing #1574 and #1571.

Replaced strided_dataset with padded_dataset class. Added support all the way up to CAGRA code.

Old class structure (Classes + Inheritance):

New Class Structure (ContainerType Tags + Composition):

Inheritance is removed entirely and all dataset types are on the same level of the inheritance tree.

3 Levels:

Storage

Container tags

Public dataset/view aliases

Ownership

The index and cagra::build / cagra::index do not own raw vector storage, they only take views.

callers (or the C merged holder) must keep backing memory alive for as long as the index is used.

The old code had a type-erased std::unique_ptr<dataset_view<...>>, i.e. non-owning view handles. The new code uses templates on the index type which determines the type of dataset_view the index holds.

ACE v.s. non-ACE paths on Host

ACE path copies datasets that can't entirely fit in CPU memory in chunks onto GPU memory by calling make_padded_dataset. This is 1x memory on CPU and 1x memory on GPU.

Return types:

Used mainly to maintain lifetime of dataset.

cuvs_cagra_c_api_lifetime_holder

unique_ptr<vpq_dataset> vpq_owner
unique_ptr padded_dataset_owner
raft::device_matrix dataset
cagra::index idx
It is a single C++ struct in cagra.cpp that groups the real cagra::index with any extra heap-owned things the C API had to create so the index’s non-owning views stay valid.

Miscellaneous: Extend Serialize Deserialize

Will fill in later

Factories:

make_device_padded_dataset_view
make_host_padded_dataset_view
make_device_padded_dataset
make_host_padded_dataset
make_vpq_dataset
- in pq.hpp and pq.cu
make_merged_dataset
- in cagra.hpp

Places where make_padded_dataset/view are called internally (not by user):

Host non-ACE path

cpp/src/neighbors/cagra_build_inst.cu.in
cagra_from_host_padded in cpp/src/neighbors/iface/iface.hpp
c/src/neighbors/cagra.cpp

Tiered CAGRA

update_cagra_ann_dataset_for_stride
build_upstream_ann

Ownership in Downstream Functions:

build() takes dataset_view only.
Downstream functions search / serialize / deserialize / merge only take views.

Improvements:

build() function previously only supported device dataset inputs. It now supports host dataset inputs.

Breaking Changes for Dataset API:

The following functions are removed since index no longer owns the dataset, index only takes views:

Removed all owning dataset based builds. Build only takes views.
Removed all update_dataset() overloads that take owning dataset. Update_dataset() only takes views.
Removed old functions that took mdspan or derivatives of mdspan.

4 cases where index previously owned dataset [all deprecated paths]:

2 edge case build() paths when attach_dataset_on_build == true and a successful dense attach:

Non-ACE / typical padded attach: rows live under index_owning_dataset_storage_ (type-erased owning wrapper, commonly device_padded_dataset).
ACE in-memory device_matrix attach: rows live under index_owning_dataset_storage_ (optional raw device_matrix).

Compression Param:

implicit vpq dataset creation within build() when compression==True
this forces index to own new vpq dataset which violates our new contract that we want index to never own dataset.

Merge:

MERGE path: merge() internally creates merged_dataset on a deprecated internal merged dataset creation path. Here, index takes ownership of merged_dataset by storing it in index_owning_dataset_storage_ .

These paths have since been removed.

Attach Dataset

Previously, in the old code, ACE attach_dataset_on_build called make_device_padded_dataset on host dataset which did a H2D copy in order to attach dataset to final index.
cpp/src/neighbors/detail/cagra/cagra_build.cuh
This has since been removed. attach_dataset_on_build is disabled for host build paths. This avoids a H2D copy.

Compressed Dataset

Removed old code that did compressed dataset creation within build. This should only happen in the factory.

Merged Dataset

Removed implicit memory allocation within merge(), memory allocation now delegated to make_merged_dataset() factory. Removed index ownership within merge.

Deserialize

Removed index ownership of dataset during deserialization. Now users are expected to create/declare the dataset type to be deserialized and then pass it as a reference to the deserialize() function which will then populate this dataset and return it to the caller.

Helpers

cagra_required_row_width
matrix_actual_row_width
matrix_row_width_matches_cagra_required
convert_dataset_view_to_padded_for_graph_build
convert_host_to_device_index
attach_device_dataset_on_host_index

How to attach a compressed dataset onto an uncompressed index?

construct a new compressed index
copy over the graph and other params from the uncompressed index
delete the old uncompressed index
attach the vpq dataset onto the compressed index

How to attach a searchable device dataset onto an index built with host build?

Convert host index to device index with helper function convert_host_to_device_index
a. Utilizes map of host dataset type to device dataset type counterpart

TODOs:

Bring back Host functions [DONE]
Mark any old functions that are no longer used as [DONE]
Use templates wherever possible. Shift towards composition rather than inheritance [DONE]

Recent Updates:

build_ace() and build() functions merged on Public API surface
removed build_result(), ace_build_result(), and merge_result()
deprecated internal vpq dataset creation inside build() when index_params::compression == true --> moved to make_vpq_dataset() factory
deprecated internal merged dataset creation inside merge() --> moved to make_merged_dataset() factory. For backwards compatibility, have index take ownership of deprecated internal merged dataset creation path ONLY.
build() and build_ace() both had a attach_dataset_on_build which requires ownership of dataset. Ownership is given to index temporarily. This will later be deprecated. Users will be expected to pass padded dataset on device and call search() directly. Attach_dataset_on_build will no longer be supported for host builds.
added host versions of dataset API
templated build() and downstream functions to work on host datasets
added index template type conversion helpers

Future PRs:

PR#2: Add Support for Compressed Datasets

pq_dataset
bbq_dataset
rabitq_dataset
sq_dataset

PR#3: Migrate Rest of Algorithms to use Dataset API

HNSW
IVF
Vamana
Scann
Brute Force

copy-pr-bot · 2026-02-24T23:21:25Z

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

aamijar · 2026-02-25T01:51:35Z

/ok to test 5447a4c

aamijar · 2026-02-25T02:12:45Z

/ok to test 17ab09d

achirkin · 2026-02-25T06:06:02Z

NB: I updated the label to breaking, since the description implies removal of a publicly visible class strided_dataset

cjnolet · 2026-02-25T22:22:05Z

Does the dataset(_view) type bring anything on top of mdarray/mdspan in that case?

@achirkin The problem w/ using mdspan/mdarray for this is that it's not carrying along the proper information to either the algorithms nor the user (which is why we created this specialized class for this in the first place!).

Two immediate reasons why this API is necessary:

The user should not have to know that they need to pad a dataset in order to use cagra without the additional copy. They should not need to know how any of these algorithms work internally. They should, however, need to know that CAGRA expects a padded dataset, and they should have an API to construct one so that they can own the dataset class and not have cagra creating one under the hood.
APIs, especially the graph-based APIs, should be able to accept as inputs data which has been quantized using a metod like PQ, which carries with additional information. In the case of PQ, the codebooks are needed to compute the distances. This again decouples the quantization from the algorithm (CAGRA-Q does not need to do its own quantization. It should just accept the quantized vectors). We're being asked for the same behavior with Vamana.

This new API solves both of these problems while leaving the control over the memory ownership entirely in the user's hands. We've discussed this for a long time. We've known this is needed for a long time. it's time to prioritize this and get it done. I agree that an anstract class might make more sense, but ultimately we should not be moving any owneship over to the algorithm (the user should maintain ownership over the class and underlying memory the entire time).

HowardHuang1 · 2026-03-04T23:24:04Z

The doc that outlines some of the API design choices can be found in slack. Let me know if there are any parts of the design that can be altered to better suit our users' needs.

The following files are test case files I've added and can be ignored for now. They will be removed before the final merge with upstream repo:

cagra_build_view_only.cu
cagra_padded_dataset.cu
cagra_vpq_build_result.cu
dataset_compression.cu
dataset_types.cu

cjnolet · 2026-03-25T14:54:12Z

/ok to test 70b6b58

achirkin

Sorry for being so late for the second round of reviews! I have a few suggestions for the dataset type hierarchy.

Need to get this ABI breaking change in for this release (we can't push this to 26.10). It was mentioned that we need to verify benchmarks. I completely agree with this. We will evaluate further in 26.10.

The expectation is that these changes really shouldn't affect performance, and they don't affect the search path at all. We'll demonstrate this in the follow-up changes, though.

… had a type erased index input which supported both host and device input index for serialize_to_hnswlib. HNSW search happens only on host. Host path was accidentally removed when new Dataset API was introduced and type erased index was removed. Added host index path back. This fixes a test failure which originates from C++ and C API layer but manifests only in java.

imotov · 2026-07-30T22:10:06Z

+    /**
+     * Internal wiring hook used by the Java wrapper implementation.
+     */
+    public final void setDelegate(AutoCloseable delegate, long handleAddress) {


Exposing this as a part of external API is a bit scary considering that end user should probably never ever try to change that.

imotov · 2026-07-30T22:18:26Z

+  PaddedDatasetView makePaddedDatasetView(CuVSMatrix dataset) throws Throwable;
+
+  /** Create a caller-owned standard dataset view handle from a matrix. */
+  StandardDatasetView makeStandardDatasetView(CuVSMatrix dataset) throws Throwable;


Is this used anywhere?

@aamijar @tarang-jain this is stale right? If I remember correctly, we consolidated padded and standard factories and removed views at language wrappers. Can I get confirmation on this?

@HowardHuang1, the padded and standard factories are still intentionally separate functions in the C API. We still have view functions in the C API. We can reassess whether we need views in the language wrappers (Python, Java, Go, Rust) in a follow up and can add that in the doc.

…originates on host. Merge() can no longer merge host data. Merge() only takes data on device

yan-zaretskiy · 2026-07-31T00:32:38Z

/ok to test 1618150

…tests

…-Dataset-API

… OOM. However, we should re-evaluate this because I don't think this is good for the new DtasetAPI as the index return type can now be host index when GPU hits OOM but the user still expects device index. With the new templating on the datasets and indexes, this can cause user/caller confusion.

cjnolet · 2026-07-31T00:45:40Z

/ok to test 7011fe6

…earch. TODO: need to revisit this later. This is a temporary patch to fix failing java tests and avoid exposing a public tiered index update dataset function for users. Instead we use CAGRA's existing update dataset functions and keep it tiered index's responsibility to call this internally rather than leaving the responsibility to the users

HowardHuang1 · 2026-07-31T01:32:14Z

/ok to test c727de2

Update C Tests to Use Getters

aamijar · 2026-07-31T01:54:50Z

/ok to test e28a173

cjnolet · 2026-07-31T01:58:46Z

/ok to test e28a173

… padded datasets onto host or standard indexes prior to search within Java tests.

HowardHuang1 · 2026-07-31T02:36:00Z

/ok to test 6eebea0

…ndexImpl.java Co-authored-by: Igor Motov <igor@motovs.org>

cjnolet · 2026-07-31T03:34:36Z

/ok to test eb6e2db

…d throw success

HowardHuang1 · 2026-07-31T03:47:45Z

/ok to test ae4757a

HowardHuang1 requested review from a team as code owners February 24, 2026 23:21

github-project-automation Bot added this to Unstructured Data Processing Feb 24, 2026

aamijar added non-breaking Introduces a non-breaking change feature request New feature or request labels Feb 25, 2026

aamijar assigned HowardHuang1 Feb 25, 2026

aamijar moved this to In Progress in Unstructured Data Processing Feb 25, 2026

achirkin added breaking Introduces a breaking change and removed non-breaking Introduces a non-breaking change labels Feb 25, 2026

achirkin reviewed Feb 25, 2026

View reviewed changes

Comment thread cpp/include/cuvs/neighbors/cagra.hpp

aamijar reviewed Feb 25, 2026

View reviewed changes

Comment thread cpp/CMakeLists.txt Outdated

seunghwak mentioned this pull request Feb 27, 2026

[WIP] Clarify dataset ownership and allocation semantics #1738

Closed

HowardHuang1 requested a review from a team as a code owner February 28, 2026 03:34

cjnolet mentioned this pull request Mar 10, 2026

[FEA] Implement binary dataset and accept in CAGRA #1906

Open

mohanprasand-nuvai mentioned this pull request Mar 23, 2026

Binary dataset support in CAGRA (upstream #1906) Nuvai/cuvs#5

Closed

5 tasks

HowardHuang1 changed the title ~~[WIP] New Dataset API Clarifying Ownership~~ [REVIEW] New Dataset API Clarifying Ownership Mar 25, 2026

HowardHuang1 changed the base branch from main to release/26.04 March 25, 2026 01:03

HowardHuang1 requested review from a team as code owners March 25, 2026 01:22

HowardHuang1 requested a review from bdice March 25, 2026 01:22

tfeher requested changes Apr 1, 2026

View reviewed changes

Comment thread cpp/include/cuvs/neighbors/cagra.hpp Outdated

Comment thread cpp/include/cuvs/neighbors/cagra.hpp Outdated

tfeher changed the base branch from release/26.04 to main April 1, 2026 23:31

achirkin requested changes Apr 8, 2026

View reviewed changes

tarang-jain and others added 2 commits July 30, 2026 14:41

further addr reviews

84e4058

imotov reviewed Jul 30, 2026

View reviewed changes

aamijar and others added 4 commits July 31, 2026 00:11

device to host padded C test and some refactor in cagra.cpp

5282b21

Update Java merge test call site to call makePaddedDataset when data …

913baf4

…originates on host. Merge() can no longer merge host data. Merge() only takes data on device

rust: update bindings to reflect the recent C API changes

1618150

fix c tests

c99b7c4

tarang-jain and others added 3 commits July 30, 2026 17:32

Merge branch 'c-tests' of https://github.com/tarang-jain/cuvs into c-…

ae897d8

…tests

Merge branch 'HH-Dataset-API' of github.com:HowardHuang1/cuvs into HH…

f3ae492

…-Dataset-API

tarang-jain and others added 2 commits July 30, 2026 18:45

Merge branch 'HH-Dataset-API' into c-tests

4818c93

Merge pull request #5 from tarang-jain/c-tests

e28a173

Update C Tests to Use Getters

Fix failing Java test cases. Add updateDataset calls to attach device…

6eebea0

… padded datasets onto host or standard indexes prior to search within Java tests.

HowardHuang1 and others added 2 commits July 30, 2026 20:21

Update java/cuvs-java/src/main/java22/com/nvidia/cuvs/internal/CagraI…

19be293

…ndexImpl.java Co-authored-by: Igor Motov <igor@motovs.org>

Update java/cuvs-java/src/main/java22/com/nvidia/cuvs/internal/CagraI…

eb6e2db

…ndexImpl.java Co-authored-by: Igor Motov <igor@motovs.org>

Fix failing clib tests. Change HNSWLIB to take in host index again an…

ae4757a

…d throw success

imotov mentioned this pull request Jul 31, 2026

Serialize CAGRA index with device-padded dataset layout NVIDIA/cuvs-lucene#192

Merged

cjnolet merged commit 2db820e into NVIDIA:release/26.08 Jul 31, 2026
83 of 85 checks passed

github-project-automation Bot moved this from In Progress to Done in Unstructured Data Processing Jul 31, 2026

HowardHuang1 mentioned this pull request Jul 31, 2026

New Dataset API Clarifying Ownership #2 #2391

Open

Uh oh!

Conversation

HowardHuang1 commented Feb 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overview

Old class structure (Classes + Inheritance):

New Class Structure (ContainerType Tags + Composition):

Ownership

ACE v.s. non-ACE paths on Host

Return types:

Miscellaneous: Extend Serialize Deserialize

Factories:

Places where make_padded_dataset/view are called internally (not by user):

Ownership in Downstream Functions:

Improvements:

Breaking Changes for Dataset API:

4 cases where index previously owned dataset [all deprecated paths]:

Attach Dataset

Compressed Dataset

Merged Dataset

Deserialize

Helpers

How to attach a compressed dataset onto an uncompressed index?

How to attach a searchable device dataset onto an index built with host build?

TODOs:

Recent Updates:

Future PRs:

Uh oh!

copy-pr-bot Bot commented Feb 24, 2026

Uh oh!

aamijar commented Feb 25, 2026

Uh oh!

aamijar commented Feb 25, 2026

Uh oh!

achirkin commented Feb 25, 2026

Uh oh!

Uh oh!

cjnolet commented Feb 25, 2026

Uh oh!

Uh oh!

HowardHuang1 commented Mar 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cjnolet commented Mar 25, 2026

Uh oh!

Uh oh!

Uh oh!

achirkin left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

imotov Jul 30, 2026

Choose a reason for hiding this comment

Uh oh!

imotov Jul 30, 2026

Choose a reason for hiding this comment

Uh oh!

HowardHuang1 Jul 31, 2026

Choose a reason for hiding this comment

Uh oh!

aamijar Jul 31, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

yan-zaretskiy commented Jul 31, 2026

Uh oh!

cjnolet commented Jul 31, 2026

Uh oh!

HowardHuang1 commented Jul 31, 2026

Uh oh!

aamijar commented Jul 31, 2026

Uh oh!

cjnolet commented Jul 31, 2026

Uh oh!

HowardHuang1 commented Jul 31, 2026

HowardHuang1 commented Feb 24, 2026 •

edited

Loading

HowardHuang1 commented Mar 4, 2026 •

edited

Loading