Uplift third_party/tt-metal to 9d69fb143bac50983dff914c5348539d0a7d2021 2025-01-31

[vmilosevic]

This PR uplifts the third_party/tt-metal to the 9d69fb143bac50983dff914c5348539d0a7d2021

• Replace Shape with SimpleShape after metal change 452da80

[brataTT]

Brings 42 metal changes

tt-torch CI Passing
tt-forge-fe CI Passing
tt-xla CI Passing

git log --oneline 064646adfcb5f0b923a24a633d532552bd43e06d..9d69fb143bac50983dff914c5348539d0a7d2021
9d69fb143b #16339: seperate dispatch constants (#17340)
56be6a5739 #0: Add api for fabric async write
7709078eac 17059: Update golden functions (#17133)
452da80c39 Remove deprecated get_shape() from Moreh ops (#17355)
89a06612f8 [skip ci] disable-pr-gate (#17412)
eec8d43480 #17291: Move dispatch_s queries out of IDevice interface to a dispatch query layer
9240420a7c [skip-ci] Revert the timeout bump
a44e186505 Extract CommandQueue interface (#17352)
1550e2a67b (tag: v0.56.0-rc3) [skip-ci] Debug strange error (#17381)
cc61668b17 [skip ci] Revert "#15414: Read annotation data to determine job-level failure signature and reason" (#17409)
8058f83694 #0: [skip ci] Fix broken README link for "Programming Mesh of Devices"
0983ce0122 [skip ci] Revert "#17374: Add concurrency group for _produce-data.yaml" (#17404)
db67d1f17c [skip ci] #17374: Add concurrency group for _produce-data.yaml (#17402)
dab9fe3fd8 #0: TT-Mesh bug fix MeshCQ/MeshWorkload on device indexing
2024847763 #15414: Read annotation data to determine job-level failure signature and reason (#17308)
242d4e625b [skip-ci] Bump timeout (#17397)
840c0b0e55 [tt-train] Add scatter workaround while proper version is in development (#17384)
230f71ba20 #0: Correcting bad dim check in CCL tests (#17392)
f358ebe390 Replace ttnn::Shape/LegacyShape with SimpleShape in Python (#17341)
12ea304aa0 #0: Add missing include for `work_split.hpp` (#17390)
dae387e049 #0: Fix incorrect assertion introduced in #17259 (#17386)
9f04f9c61e [skip-ci] Test for a bad state before building (#17379)
9edf9a0cac Implement ttnn.sampling op for top-k, top-p sampling (#17136)
20360c539f #17312: Fix type error when saving report config to json file (#17350)
b6947b510f #15889: Fix handling of mantissa rounding to respect ties round to even 
0856c50a23 #16149: Add DeviceCommandCalculator to calculate command size
3992e686fb #0: Use posted writes for profiler.
290d99a0a8 [skip ci]: Only produce cicd data on workflow runs that are success/failure/cancelled (ignore skipped runs) (#17371)
99302c5029 fix sub-height height sharded WH transpose by setting output memory config to width sharded (#17147)
6d559e2f79 #17215: Initial MeshBuffer integration with TTNN (#17259)
dbe37c2dcf #17322 Remove transpose cpp unit tests (#17326)
a54a52bb23 Use padded shape to construct output memory config in `ttnn.view` to fix nightly pipelines (#17366)
4c02dadf3c Make runtime_args_ptr const ref to solve clang-tidy error due to 58f9654bc (#17364)
76b5bd708f #0: [skip ci] Fix retry comparison which causes endless retries until pass (#17367)
0e61fc0fb9 #0: Revert "#16138:  W-broadcasting for sharded tensors"
cb2cbd30e0 #17134: Remove unused SD components (#17301)
bb8787568e Fix shard and physical height mismatch in `ttnn.convert_to_chw` tests
df23d32a49 Fix non-contiguous CB indices in `ttnn.convert_to_chw`
68f9add0cc #0: expand halo documentation and fix images (#16802)
4274236190 Add perf report for reduce scatter async (#17223)
eb36bff16d Fix debug checks for bank assignments when initializing the allocator
32bb8292c1 Add perf reporting for ccl async mode (#16658)

[kmabeeTT]

Thanks, Brata! 🔥