Ship the OpenVINO delegate as its own library - #21770
Open
shoumikhin wants to merge 28 commits into
Open
Conversation
Contributor
Author
shoumikhin
requested review from
kirklandsign,
larryliu0820 and
mergennachin
as code owners
August 12, 2026 06:08
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21770
Note: Links to docs will display an error until the docs builds have been completed. ❌ 30 New Failures, 2 Unrelated Failures, 30 Unclassified FailuresAs of commit 1ccd490 with merge base ed65b12 ( NEW FAILURES - The following jobs have failed:
UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:
FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
This was referenced Aug 12, 2026
shoumikhin
force-pushed
the
gh/shoumikhin/96/head
branch
from
August 12, 2026 16:19
9428da3 to
2cdf978
Compare
shoumikhin
force-pushed
the
gh/shoumikhin/105/head
branch
from
August 12, 2026 16:19
00563b2 to
d3de6d5
Compare
## The problem The wheel can carry the CUDA delegate, but no published wheel contains one: there is no CUDA row in any workflow, so a GPU user has to build from source. ## The change Add the workflows that build and publish CUDA wheels for Linux x86_64 and aarch64, and a smoke test that checks each wheel from the artifact itself. The build machines for these rows have no GPU, so the smoke test does not execute a model; it verifies the CUDA libraries are present, that the declared runtime matches the wheel's CUDA version, that nothing resolves through the build machine's toolkit, and that the shipped device code covers every GPU architecture the row claims. ``` executorch-1.5.0-cp312-cp312-manylinux_2_28_x86_64.whl +cu130 ``` A release publishes CUDA 12.6, 13.0 and 13.2, for Python 3.10 through 3.13. A pull request builds a single row instead of all twelve, because a full matrix costs hours for little extra signal. Which GPU architectures each row compiles for is chosen per row rather than detected on the builder. Detecting it would produce a wheel carrying device code for whatever machine happened to build it, which installs fine and then fails at the first GPU call. The aarch64 CUDA 12.6 row also compiles for compute capability 8.7, which is an embedded module. Every other row lists only the architectures the published PyTorch build for that train covers, and by that rule 8.7 would be left out, because the generic aarch64 build of this train carries 8.0 and 9.0 only. It is included because this is the only row whose CUDA major version matches what that module's software release ships, and because this wheel declares no PyTorch dependency: a user there supplies the build that carries their architecture. Leaving 8.7 out does not protect them from a bad pairing, it only removes the device code they need. Without it, a model reaching one of the shipped optional operators, quantized matrix multiply, sort or random number generation, fails at the first launch on that device. Two guards keep a release honest: - if the shared matrix generator stops offering a combination this policy advertises, the step fails instead of quietly publishing fewer wheels. A missing job is otherwise a green check for a wheel that was never built. - if a row reaches the architecture list with no CUDA version, the build refuses rather than falling back to the builder's GPU. A TORCH_CUDA_ARCH_LIST that holds only named GPU families PyTorch accepts, such as "Hopper", now fails to configure instead of quietly leaving CMAKE_CUDA_ARCHITECTURES unset and taking the compiler default. The three named forms CMake itself understands, "native", "all" and "all-major", are rejected before this logic runs: torch resolves the list with its own bundled CUDA architecture module, which does not know those names and stops the configure. That is upstream behaviour, not something this change introduces or can work around, so a caller has to name architectures explicitly. Windows CUDA is deliberately absent. The separate shared libraries this wheel exists to ship are Linux only today, so a Windows CUDA wheel would carry a delegate a C++ application still could not link. ## Test plan - built the full release matrix, twelve wheels, and confirmed each one's contents match the row it claims: the CUDA libraries present, the CUDA runtime declared, and device code for every GPU architecture the row advertises. - ran a GPU model end to end from a CI-built wheel on three NVIDIA GPUs covering three device architectures, with output identical to eager PyTorch on each (largest absolute difference 0), and inspected the wheel for a fourth device it cannot execute on. - ran the matrix filter over generated inputs, including incomplete and malformed ones, and confirmed it refuses rather than publishing a partial release: a missing CUDA version, a missing python, or a python present on rows this policy does not build are each reported by name. - confirmed a CPU row still produces a CPU wheel on a builder that happens to have a CUDA toolkit installed. - the newest architecture also ships in its portable form, so a GPU newer than any in the row can still run by having the driver compile it at load time. Checked with `cuobjdump --list-ptx`, since `--list-elf` prints identical output whether or not the portable form is present. - every library that carries GPU device code covers the whole row on its own. - the declared CUDA packages are compared against the expected set in BOTH directions. A one-way comparison accepted a wheel that omitted required packages, and a name-suffix comparison accepted cross-train names because for CUDA 13 the suffix is empty. - the python axis is an allowlist, matching the CUDA axis. Testing only the disabled list let any python not on it through: a 3.9 row was emitted successfully. - `install_utils.py` is in both CUDA workflows' path filters. It owns the supported CUDA train list and the toolkit detection, so a change there previously ran no CUDA wheel job. - requesting the JetPack rows fails with its own reason instead of the generic empty-matrix message, since both of its lists are deliberately empty and no workflow asks for them. - torchao keeps its CUDA channel where that channel exists. Falling back to the plain nightly index was needed only on aarch64, where the CUDA channel publishes nothing, and doing it everywhere changed which torchao an x86_64 install resolves. - the CUDA smoke test now asserts the QnnBackend and OpenvinoBackend registrations that a CPU Linux row asserts. The CUDA build enables OpenVINO on every Linux architecture and downloads the QNN SDK on x86_64, so a CUDA wheel carries both backends; a previous premise that "a CUDA row is not built with them" was false, and dropping the checks meant those two backends were unverified on every CUDA wheel. Known gap: no automated job runs a CUDA model on real hardware before publication. Running a model on real hardware is a separate release-time step that a person owns today, not an automated job wired into these workflows. ghstack-source-id: 95d67c7 ghstack-comment-id: 5220374521 Pull-Request: #21668
The OpenVINO delegate was compiled into the Python extension, so only Python could
use it. A C++ application had no way to link it from the wheel, which is the same gap
the other delegates had before they were split out.
It now builds as its own library and ships beside the runtime, so an application can
ask for it:
find_package(executorch REQUIRED COMPONENTS backend_openvino)
target_link_libraries(my_app PRIVATE executorch::backend_openvino)
Only the adapter ships here. The OpenVINO runtime itself is loaded at run time and is
not part of the wheel, so the component's documentation says to install it separately,
and the wheel does not grow meaningfully: the adapter is a few kilobytes.
The library resolves the runtime from the shared runtime library rather than from the
static core, so the delegate registers into the one registry the process has instead of
a second private one.
Test Plan: built a wheel with OpenVINO enabled on Linux x86_64 and inspected it.
The library ships at executorch/lib/libexecutorch_backend_openvino.so. Its delegate
symbols moved out of the Python extension, 13 in the library and 0 in the extension,
so the process has one copy and one registration. Its only ExecuTorch dependency is
the shared runtime library, and it records no link dependency on OpenVINO, confirming
the runtime is still loaded at run time. Added an ownership row to the shared library
test so a future change that puts the delegate back into the extension fails there.
ghstack-source-id: b6b210a
ghstack-comment-id: 5263017453
Pull-Request: #21770
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The OpenVINO delegate was compiled into the Python extension, so only Python could
use it. A C++ application had no way to link it from the wheel, which is the same gap
the other delegates had before they were split out.
It now builds as its own library and ships beside the runtime, so an application can
ask for it:
Only the adapter ships here. The OpenVINO runtime itself is loaded at run time and is
not part of the wheel, so the component's documentation says to install it separately,
and the wheel does not grow meaningfully: the adapter is a few kilobytes.
The library resolves the runtime from the shared runtime library rather than from the
static core, so the delegate registers into the one registry the process has instead of
a second private one.
Test Plan: built a wheel with OpenVINO enabled on Linux x86_64 and inspected it.
The library ships at executorch/lib/libexecutorch_backend_openvino.so. Its delegate
symbols moved out of the Python extension, 13 in the library and 0 in the extension,
so the process has one copy and one registration. Its only ExecuTorch dependency is
the shared runtime library, and it records no link dependency on OpenVINO, confirming
the runtime is still loaded at run time. Added an ownership row to the shared library
test so a future change that puts the delegate back into the extension fails there.