feat: support ascend and nvidia accelerators with runc - #45
Open
mhsong1998-dot wants to merge 1 commit into
Open
mhsong1998-dot wants to merge 1 commit into
mhsong1998-dot wants to merge 1 commit into
Conversation
mhsong1998-dot
force-pushed
the
feat/ascend-npu-runc
branch
3 times, most recently
from
September 2, 2026 02:37
5d324b9 to
49108b7
Compare
mhsong1998-dot
changed the title
mhsong1998-dot
force-pushed
the
feat/ascend-npu-runc
branch
2 times, most recently
from
September 3, 2026 02:24
73742d2 to
fe7ea77
Compare
mhsong1998-dot
marked this pull request as ready for review
September 4, 2026 03:53
mhsong1998-dot
force-pushed
the
feat/ascend-npu-runc
branch
from
September 4, 2026 04:01
fe7ea77 to
eb9e16b
Compare
mhsong1998-dot
force-pushed
the
feat/ascend-npu-runc
branch
3 times, most recently
from
September 9, 2026 09:31
9b5415f to
0df9995
Compare
Package optional native runc and accelerator support across node images, Helm, Terraform, standalone deployment, and the Python SDK. Support NVIDIA GPUs with runsc or runc and physical Ascend NPUs with runc. Install Ascend integration from a checksum-pinned standalone adapter release while keeping the trusted mount profile in sandboxd. Update sandboxd to the matching external-adapter protocol implementation. Signed-off-by: mhsong2 <songminhui2@huawei.com>
mhsong1998-dot
force-pushed
the
feat/ascend-npu-runc
branch
from
September 14, 2026 10:43
0df9995 to
8d6cf55
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Package optional native runc and accelerator support across the node image, Helm, Terraform, standalone deployment, and Python SDK. NVIDIA GPUs can use runsc or runc; physical Ascend NPUs use runc.
The node image no longer builds the Ascend adapter from sandboxd or reads MindCluster source. When Ascend is enabled, it downloads the separately released ascend-oci-adapter v0.1.0-rc.2, verifies its pinned SHA-256, and installs the binary and redistribution licenses. Version, URL, and checksum remain jointly overridable for a future release. sandboxd's mount profile remains the trusted allowlist.
This branch is rebased on current AKernel main and pins sandboxd commit
2eb7f0d21eb446b06ead6b898f56bbbec94b66c9. The final PR history is one signed commit.Validation at AKernel head
8d6cf55b03499620327d75145fd3b4f37ac5e940:virtiofsdstage and the Ascend adapter stages retainedabe3b926750c2467dc5544645bae530ec4bac8572615b0678892a4408d26954c, contained the required binary and licenses, and reported schema 1 with provider versionv0.1.0-rc.2+mindcluster-v26.2.0.beta.1go test ./...in cn-north-4 CCE with Go 1.25.5 on Linux amd64npu-smi, ACL runtime initialization and memory transfers, exclusive allocation, release, device reuse, and a sandboxed CANN Add result of[11, 22, 33, 44]Depends on inclusionAI/sandboxd#39. Adapter source review and release: openYuanrong-mirror/ascend-oci-adapter#1.