[WIP][DRAFT] Use MeshDevice 1x1 instead of Device #18470
Draft
Conversation
This file contains 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
…d add to pipeline select (#18348) ### Ticket #17496 ### Problem description We did not have TG nightly tests as a selectable pipeline in choose your pipeline. This was inconvenient for developers. ### What's changed * Split out nightly tests workflow into wrapper + impl * Add impl to choose your pipeline ### Checklist - [ ] [All post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/all-post-commit-workflows.yaml) CI passes - [ ] [Blackhole Post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/blackhole-post-commit.yaml) CI passes (if applicable) - [ ] [Model regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-models.yaml) CI passes (if applicable) - [ ] [Device performance regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-device-models.yaml) CI passes (if applicable) - [ ] **(For models and ops writers)** Full [new models tests](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml) CI passes (if applicable) - [ ] New/Existing tests provide coverage for changes --------- Co-authored-by: Andrew Fuller <[email protected]>
### Ticket N/A ### Problem description Expose mesh events to TTNN, further integration of TT-distributed with TTNN. ### What's changed Related cleanups bundled in this PR: * Make `EnqueueRecordEvent`, `EnqueueRecordEventToHost` return `MeshEvent` by value, instead of accepting `std::shared_ptr<MeshEvent>` and mutating the reference internally. * Make `EnqueueWaitForEvent` and `EventSynchronize` accept `MeshEvent` by constant reference. * Expose `MeshCoordinateRange` to TTNN - this is needed for the `MeshEvent` APIs. ### Checklist - [X] [All post commit](https://github.com/tenstorrent/tt-metal/actions/runs/13577312078) - [X] New/Existing tests provide coverage for changes - ran `MeshEvents` tests from `distributed_unit_tests`
### Problem description It is not easy to customize different neural networks. ### What's changed Now all created NNs are immutable by desing. Added operator () to the base class. All should derive. Id it is possible you should always store std::shared_ptr<ModuleBase> instead of your class. ### Checklist - [x] [All post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/all-post-commit-workflows.yaml) CI passes - [x] New/Existing tests provide coverage for changes
… into sminakov/all-mesh2
This reverts commit 351edaf.
This reverts commit 01ca8ac.
…17790) ### Ticket Link to Github Issue #15136 ### Problem description Limited parallelization for tilize/untilize along one dimension, which affects the performance ### What's changed This PR adds parallelization along last two dims for tilize/untilize with padding. For large tensors, the operations use more cores and improve the perf by around 30 times for the tests added. The average device samples/s is more than 3.5 times larger for models like vgg11 and vgg16. It is also improved for some Bert tiny tests. ### Checklist - [x] [All post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/all-post-commit-workflows.yaml) CI passes https://github.com/tenstorrent/tt-metal/actions/runs/13243407566 - [ ] [Blackhole Post commit](https://github.com/tenstorrent/tt-metal/actions/workflows/blackhole-post-commit.yaml) CI passes (if applicable) - [ ] [Model regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-models.yaml) CI passes (if applicable) - [x] [Device performance regression](https://github.com/tenstorrent/tt-metal/actions/workflows/perf-device-models.yaml) CI passes (if applicable) https://github.com/tenstorrent/tt-metal/actions/runs/13249354224 - [ ] **(For models and ops writers)** Full [new models tests](https://github.com/tenstorrent/tt-metal/actions/workflows/full-new-models-suite.yaml) CI passes (if applicable) - [ ] New/Existing tests provide coverage for changes
This change updates the sliding window config tensor generation to use structs instead of tuples to improve readibility and type safety.
### Ticket None ### Problem description A dependency on ASIO was introduced in #18395, but we only "accidentally" have it available because it (sometimes) comes along with other Boost headers. CI was fine because it happened to be there. But in some envs it isn't there (or too old). ### What's changed Declared our dependency on it.
This is in addition to the existing LICENSE file to help clarify Tenstorrent's position and understanding of specific aspects of the existing LICENSE Signed-off-by: John 'Warthog9' Hawley <[email protected]>
… into sminakov/all-mesh2
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Ticket
Link to Github Issue
Problem description
Provide context for the problem.
What's changed
Describe the approach used to solve the problem.
Summarize the changes made and its impact.
Checklist