Metadata-Version: 2.4
Name: tensorplay-mlu
Version: 0.1.0
Summary: Cambricon MLU device backend for TensorPlay
License:                                  Apache License
                                   Version 2.0, January 2004
                                http://www.apache.org/licenses/
        
           TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
        
           1. Definitions.
        
              "License" shall mean the terms and conditions for use, reproduction,
              and distribution as defined by Sections 1 through 9 of this document.
        
              "Licensor" shall mean the copyright owner or entity authorized by
              the copyright owner that is granting the License.
        
              "Legal Entity" shall mean the union of the acting entity and all
              other entities that control, are controlled by, or are under common
              control with that entity. For the purposes of this definition,
              "control" means (i) the power, direct or indirect, to cause the
              direction or management of such entity, whether by contract or
              otherwise, or (ii) ownership of fifty percent (50%) or more of the
              outstanding shares, or (iii) beneficial ownership of such entity.
        
              "You" (or "Your") shall mean an individual or Legal Entity
              exercising permissions granted by this License.
        
              "Source" form shall mean the preferred form for making modifications,
              including but not limited to software source code, documentation
              source, and configuration files.
        
              "Object" form shall mean any form resulting from mechanical
              transformation or translation of a Source form, including but
              not limited to compiled object code, generated documentation,
              and conversions to other media types.
        
              "Work" shall mean the work of authorship, whether in Source or
              Object form, made available under the License, as indicated by a
              copyright notice that is included in or attached to the work
              (an example is provided in the Appendix below).
        
              "Derivative Works" shall mean any work, whether in Source or Object
              form, that is based on (or derived from) the Work and for which the
              editorial revisions, annotations, elaborations, or other modifications
              represent, as a whole, an original work of authorship. For the purposes
              of this License, Derivative Works shall not include works that remain
              separable from, or merely link (or bind by name) to the interfaces of,
              the Work and Derivative Works thereof.
        
              "Contribution" shall mean any work of authorship, including
              the original version of the Work and any modifications or additions
              to that Work or Derivative Works thereof, that is intentionally
              submitted to Licensor for inclusion in the Work by the copyright owner
              or by an individual or Legal Entity authorized to submit on behalf of
              the copyright owner. For the purposes of this definition, "submitted"
              means any form of electronic, verbal, or written communication sent
              to the Licensor or its representatives, including but not limited to
              communication on electronic mailing lists, source code control systems,
              and issue tracking systems that are managed by, or on behalf of, the
              Licensor for the purpose of discussing and improving the Work, but
              excluding communication that is conspicuously marked or otherwise
              designated in writing by the copyright owner as "Not a Contribution."
        
              "Contributor" shall mean Licensor and any individual or Legal Entity
              on behalf of whom a Contribution has been received by Licensor and
              subsequently incorporated within the Work.
        
           2. Grant of Copyright License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              copyright license to reproduce, prepare Derivative Works of,
              publicly display, publicly perform, sublicense, and distribute the
              Work and such Derivative Works in Source or Object form.
        
           3. Grant of Patent License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              (except as stated in this section) patent license to make, have made,
              use, offer to sell, sell, import, and otherwise transfer the Work,
              where such license applies only to those patent claims licensable
              by such Contributor that are necessarily infringed by their
              Contribution(s) alone or by combination of their Contribution(s)
              with the Work to which such Contribution(s) was submitted. If You
              institute patent litigation against any entity (including a
              cross-claim or counterclaim in a lawsuit) alleging that the Work
              or a Contribution incorporated within the Work constitutes direct
              or contributory patent infringement, then any patent licenses
              granted to You under this License for that Work shall terminate
              as of the date such litigation is filed.
        
           4. Redistribution. You may reproduce and distribute copies of the
              Work or Derivative Works thereof in any medium, with or without
              modifications, and in Source or Object form, provided that You
              meet the following conditions:
        
              (a) You must give any other recipients of the Work or
                  Derivative Works a copy of this License; and
        
              (b) You must cause any modified files to carry prominent notices
                  stating that You changed the files; and
        
              (c) You must retain, in the Source form of any Derivative Works
                  that You distribute, all copyright, patent, trademark, and
                  attribution notices from the Source form of the Work,
                  excluding those notices that do not pertain to any part of
                  the Derivative Works; and
        
              (d) If the Work includes a "NOTICE" text file as part of its
                  distribution, then any Derivative Works that You distribute must
                  include a readable copy of the attribution notices contained
                  within such NOTICE file, excluding those notices that do not
                  pertain to any part of the Derivative Works, in at least one
                  of the following places: within a NOTICE text file distributed
                  as part of the Derivative Works; within the Source form or
                  documentation, if provided along with the Derivative Works; or,
                  within a display generated by the Derivative Works, if and
                  wherever such third-party notices normally appear. The contents
                  of the NOTICE file are for informational purposes only and
                  do not modify the License. You may add Your own attribution
                  notices within Derivative Works that You distribute, alongside
                  or as an addendum to the NOTICE text from the Work, provided
                  that such additional attribution notices cannot be construed
                  as modifying the License.
        
              You may add Your own copyright statement to Your modifications and
              may provide additional or different license terms and conditions
              for use, reproduction, or distribution of Your modifications, or
              for any such Derivative Works as a whole, provided Your use,
              reproduction, and distribution of the Work otherwise complies with
              the conditions stated in this License.
        
           5. Submission of Contributions. Unless You explicitly state otherwise,
              any Contribution intentionally submitted for inclusion in the Work
              by You to the Licensor shall be under the terms and conditions of
              this License, without any additional terms or conditions.
              Notwithstanding the above, nothing herein shall supersede or modify
              the terms of any separate license agreement you may have executed
              with Licensor regarding such Contributions.
        
           6. Trademarks. This License does not grant permission to use the trade
              names, trademarks, service marks, or product names of the Licensor,
              except as required for reasonable and customary use in describing the
              origin of the Work and reproducing the content of the NOTICE file.
        
           7. Disclaimer of Warranty. Unless required by applicable law or
              agreed to in writing, Licensor provides the Work (and each
              Contributor provides its Contributions) on an "AS IS" BASIS,
              WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
              implied, including, without limitation, any warranties or conditions
              of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
              PARTICULAR PURPOSE. You are solely responsible for determining the
              appropriateness of using or redistributing the Work and assume any
              risks associated with Your exercise of permissions under this License.
        
           8. Limitation of Liability. In no event and under no legal theory,
              whether in tort (including negligence), contract, or otherwise,
              unless required by applicable law (such as deliberate and grossly
              negligent acts) or agreed to in writing, shall any Contributor be
              liable to You for damages, including any direct, indirect, special,
              incidental, or consequential damages of any character arising as a
              result of this License or out of the use or inability to use the
              Work (including but not limited to damages for loss of goodwill,
              work stoppage, computer failure or malfunction, or any and all
              other commercial damages or losses), even if such Contributor
              has been advised of the possibility of such damages.
        
           9. Accepting Warranty or Additional Liability. While redistributing
              the Work or Derivative Works thereof, You may choose to offer,
              and charge a fee for, acceptance of support, warranty, indemnity,
              or other liability obligations and/or rights consistent with this
              License. However, in accepting such obligations, You may act only
              on Your own behalf and on Your sole responsibility, not on behalf
              of any other Contributor, and only if You agree to indemnify,
              defend, and hold each Contributor harmless for any liability
              incurred by, or claims asserted against, such Contributor by reason
              of your accepting any such warranty or additional liability.
        
           END OF TERMS AND CONDITIONS
        
           APPENDIX: How to apply the Apache License to your work.
        
              To apply the Apache License to your work, attach the following
              boilerplate notice, with the fields enclosed by brackets "[]"
              replaced with your own identifying information. (Don't include
              the brackets!)  The text should be enclosed in the appropriate
              comment syntax for the file format. We also recommend that a
              file or class name and description of purpose be included on the
              same "printed page" as the copyright notice for easier
              identification within third-party archives.
        
           Copyright [2025] [zlx]
        
           Licensed under the Apache License, Version 2.0 (the "License");
           you may not use this file except in compliance with the License.
           You may obtain a copy of the License at
        
               http://www.apache.org/licenses/LICENSE-2.0
        
           Unless required by applicable law or agreed to in writing, software
           distributed under the License is distributed on an "AS IS" BASIS,
           WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
           See the License for the specific language governing permissions and
           limitations under the License.
Project-URL: Homepage, https://github.com/ohtensorplay/tensorplay-mlu
Project-URL: Repository, https://github.com/ohtensorplay/tensorplay-mlu
Project-URL: TensorPlay, https://github.com/lexing-2026/TensorPlay
Project-URL: Documentation, https://github.com/lexing-2026/TensorPlay/blob/main/docs/source/notes/backends.md
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: POSIX :: Linux
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: C++
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: tensorplay
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Dynamic: license-file

<p align="center">
  <picture>
    <source media="(prefers-color-scheme: dark)" srcset="docs/images/cambricon-lockup-white.png">
    <source media="(prefers-color-scheme: light)" srcset="docs/images/cambricon-lockup.png">
    <img src="docs/images/cambricon-lockup.png" alt="Cambricon" width="420">
  </picture>
</p>

<h1 align="center">tensorplay-mlu</h1>

**Cambricon MLU backend for [TensorPlay]** — a native, out-of-tree device
backend that makes TensorPlay run on Cambricon MLU accelerators through the
CNRT driver and the CNNL kernel library.

```python
import tensorplay as tp
import tensorplay_mlu            # compiles the native part on first load

x = tp.randn(64, 64, device="mlu")
y = (x @ x).relu().sum()
print(y.item())
```

[![License](https://img.shields.io/badge/license-Apache--2.0-blue.svg)](LICENSE)
[![Python](https://img.shields.io/badge/python-3.10%2B-blue)](https://www.python.org/)
[![TensorPlay](https://img.shields.io/badge/powered%20by-TensorPlay-orange)](https://github.com/lexing-2026/TensorPlay)

<p align="center"><a href="./README.md"><strong>English</strong></a> · <a href="./README.zh.md">中文</a></p>

---

## What this is

TensorPlay reserves one out-of-tree device slot (`PrivateUse1`) for
accelerators that live outside the framework.  This package fills that slot
for Cambricon MLU:

- memory comes from CNRT through a **caching allocator**;
- operators run on **CNNL** kernels;
- every operator without an MLU kernel yet runs on the **host** through the
  generic fallback, so models work end to end from day one;
- the device is spelled `"mlu"` in TensorPlay (`tp.ones(3, device="mlu")`).

The native sources ship inside the wheel and are compiled against the
installed Neuware toolkit the first time the backend is loaded, then cached
under `~/.cache/tensorplay_extensions`.

The same package ships a **MagicMind graph backend**: `tensorplay.compile(fn,
backend="magicmind")` lowers a captured region through Cambricon MagicMind and
runs it on the MLU — see [Graph compilation (MagicMind)](#graph-compilation-magicmind).

## Installation

### Requirements

- a Cambricon MLU card with the **Neuware toolkit** installed
  (`CNToolkit`, `CNNL`, ... — see the Cambricon user manuals)
- TensorPlay installed in the same environment (`pip install tensorplay`)
- the **MagicMind** SDK, only for the `magicmind` graph backend

### Environment variables

```bash
export NEUWARE_HOME=/usr/local/neuware   # Neuware toolkit install path
export LD_LIBRARY_PATH=$NEUWARE_HOME/lib64:$LD_LIBRARY_PATH
```

### Install the package

```bash
git clone https://github.com/ohtensorplay/tensorplay-mlu.git
cd tensorplay-mlu
pip install .            # or: pip install -e .
```

`import tensorplay` auto-loads the backend through the `tensorplay.backends`
entry point when `$NEUWARE_HOME` holds the toolkit.  You can also load it
explicitly:

```python
import tensorplay_mlu
tensorplay_mlu.load()
```

### Verify

```python
import tensorplay as tp
import tensorplay_mlu

tensorplay_mlu.load()
a = tp.randn(2, 3, device="mlu")
print((a.abs().sum()).cpu().item())   # needs an MLU server
```

## Quick start

```python
import tensorplay as tp
import tensorplay_mlu

tensorplay_mlu.load()
tp.manual_seed(0)

x = tp.randn(256, 256, device="mlu")
w = tp.randn(256, 256, device="mlu")

y = x @ w                # native CNNL matmul
z = y.relu().softmax(-1) # activations and reductions
loss = z.sum()
loss.backward()          # autograd works through the backend

print(z.cpu().shape, float(loss.cpu()))
```

## Supported operations

The backend implements the hot paths natively on CNNL:

| Area | Operators |
| --- | --- |
| Memory | `empty`, `*_like`, `copy_` (host/device, casts, broadcasting), `fill_`, `clone`, `item` |
| Binary | `add`/`sub`/`mul`/`div` with broadcasting, in-place and scalar forms |
| Unary | `exp`, `log`, `sqrt`, `pow`, activations (`relu`, `gelu`, `sigmoid`, ...) and their backward kernels |
| Blas | `mm`, `addmm`, `bmm`, `matmul`, `linear` (transposes read in place, cached algorithms) |
| Reduce | `sum`, `mean`, `amax`/`amin`, `softmax`/`log_softmax` with backward |
| Distributed | `all_reduce`, `broadcast`, `reduce`, `all_gather`, `reduce_scatter`, `all_to_all`, `gather`/`scatter`, `send`/`recv`, `barrier` on the `cncl` backend |
| Profiling | per-operation device time, kernel-level records through CNPAPI |

Everything else runs on the host through
`cpu_fallback::register_host_fallbacks`, so **every model runs**; data types
CNNL does not compute in (integers, `float64`) take the host path as well.

## Distributed (CNCL)

`dist.init_process_group("cncl")` selects a process group that runs the
collectives over Cambricon **CNCL** on MLU tensors.  The backend registers
itself when the extension loads, so no extra import or environment variable is
needed:

```python
import tensorplay as tp
import tensorplay_mlu

tensorplay_mlu.load()
import tensorplay.distributed as dist

dist.init_process_group("cncl", init_method="env://")
x = tp.ones(4, device="mlu") * dist.get_rank()
dist.all_reduce(x, op=dist.ReduceOp.SUM)
dist.destroy_process_group()
```

Launch one process per card (`torchrun`-style `RANK` / `WORLD_SIZE` /
`MASTER_ADDR` / `MASTER_PORT`), or use `tp.distributed.run`.  `AVG` is realized
as a `SUM` followed by a scale, and `PRODUCT` reductions follow CNCL's
conventions.  `monitored_barrier` and receiving from any source are not
available and raise.

## Graph compilation (MagicMind)

`tensorplay.compile(fn, backend="magicmind")` lowers a captured region
through Cambricon **MagicMind** and runs it on the MLU.  The region is
exported to ONNX, parsed into a MagicMind network, and built into an MLU
model that is cached under `~/.cache/tensorplay-magicmind`.  The cache key
covers the graph, the sample shapes and dtypes, and the settings, so a
rebuild only happens when something that could change the binary changed.

```python
import tensorplay as tp
import tensorplay_mlu

tensorplay_mlu.load()

def region(x, w, bias):
    return tp.softmax(tp.relu(x @ w + bias), 1)

f = tp.compile(region, backend="magicmind")
x = tp.randn(8, 16, device="mlu")
w = tp.randn(16, 32, device="mlu")
b = tp.randn(32, device="mlu")
print(f(x, w, b).device)          # mlu
```

Options go through `options=` (or as flat keywords):

| Option | Meaning |
| --- | --- |
| `precision` | `"fp32"` (default) or `"fp16"` |
| `mlu_archs` | architectures the built binary targets, for example `["mtp_372"]`; unset lets the builder choose |
| `cache_dir` | directory for serialized models; `None` disables the disk cache |
| `pass_through_build_failures` | raise on a build failure instead of running the region uncompiled |

MagicMind is an optional dependency.  Without it the backend is reported as
unavailable and naming it in `tensorplay.compile` explains what to install.  A
region MagicMind cannot build stays on its interpreter, so a coverage gap
never fails the whole compile.

## Performance

`benchmark/bench_mlu_vs_torch.py` in the TensorPlay repository runs this
backend against `torch + torch_mlu` on the same card, one framework per
process, identical workloads.  The chart below shows the ratio on a
Cambricon MLU card (MLU590 series, TF32 on both sides, median of 50 timed
iterations); values below 1.00 mean TensorPlay is faster:

<p align="center">
  <picture>
    <source media="(prefers-color-scheme: dark)" srcset="docs/images/benchmark-ratio-dark.png">
    <source media="(prefers-color-scheme: light)" srcset="docs/images/benchmark-ratio.png">
    <img src="docs/images/benchmark-ratio.png" alt="tensorplay-mlu vs torch_mlu benchmark ratio" width="720">
  </picture>
</p>

Representative numbers:

| Case | torch ms | tensorplay ms | tp/torch |
| --- | ---: | ---: | ---: |
| mm 4096 fp32 | 1.72 | 1.71 | 0.99 |
| mm 1024 fp32 | 0.07 | 0.07 | 0.91 |
| add 4096² fp32 | 0.38 | 0.38 | 1.00 |
| relu 4096² fp32 | 0.15 | 0.14 | 0.97 |
| gelu 4096² fp32 | 0.21 | 0.20 | 0.96 |
| softmax 4096² fp32 | 0.21 | 0.20 | 0.97 |
| sum_all 4096² fp32 | 0.10 | 0.08 | 0.88 |
| bmm 32×512×64×512 fp32 | 0.07 | 0.07 | 1.00 |
| h2d 64 MiB | 5.10 | 5.03 | 0.99 |
| d2h 64 MiB | 26.7 | 5.6 | 0.21 |
| mlp train step fp32 | 0.82 | 0.50 | 0.60 |

Elementwise, reduction and memory operators are as fast as or faster than
`torch_mlu`; host→device copies match it and **device→host copies are about
5× faster**.  Full results are reproducible with:

```bash
cd <tensorplay-repo>
BENCH_FP32_PRECISION=high python benchmark/bench_mlu_vs_torch.py --per-case
```

> **Note on TF32.** `torch_mlu` always runs float32 matmuls in TF32 — its
> `allow_tf32` setter is a no-op in the shipped build.  The benchmark
> defaults to TF32 on both sides (`BENCH_FP32_PRECISION=high`) for a
> like-for-like comparison; set `BENCH_FP32_PRECISION=highest` to force full
> float32 on the TensorPlay side.

### Distributed scaling

`benchmark/bench_mlu_distributed.py` in the TensorPlay repository runs one
data-parallel training step (a 1024→4096→1024 MLP, gradients averaged with
`all_reduce`) and reports the throughput as cards are added.  Two MLU cards,
one process each, interconnected over the host network (no InfiniBand
available), with the same per-rank batch on one and two cards:

| per-rank batch | cards | step ms | samples/s | speedup |
| ---: | ---: | ---: | ---: | ---: |
| 512 | 1 | 1.11 | 463k | — |
| 512 | 2 | 12.3 | 83k | 0.18× |
| 4096 | 1 | 5.48 | 748k | — |
| 4096 | 2 | 16.6 | 492k | 0.66× |
| 16384 | 1 | 20.9 | 783k | — |
| 16384 | 2 | 32.1 | 1.02M | 1.30× |

The crossover is set by the gradient exchange: this MLP has about 8.4M
parameters, so every step all-reduces roughly 33 MB, which costs around
10 ms over this interconnect regardless of the batch size.  Below a few
thousand samples per rank the exchange dominates and the second card only
slows the step down; from roughly 16k samples per rank the step is again
compute-bound and two cards reach 1.30× (65% of the ideal 2×).  A faster
interconnect moves the crossover to smaller batches.  One-card
throughput alone saturates near 0.78M samples/s once the batch clears
about 16k.

Reproduce with one process per card:

```bash
RANK=0 WORLD_SIZE=2 MASTER_ADDR=<rank-0-host> MASTER_PORT=<free-port> \
  python benchmark/bench_mlu_distributed.py --global-batch 32768 --json out.json
```

## Generic product names

CNNL selects its kernels and tiling tables by the product name the driver
reports.  Virtualized and rebranded cards can report a generic name (`mlu`;
CNNL logs "device name too short"), and CNNL then falls back to an
unknown-device configuration whose kernels fault the queue or return wrong
data without an error status: matrix products with 2 to 256 rows or above
2048 on a side, long reductions along the last axis, whole-tensor reductions
over 65536 elements, half precision elementwise operators above 32 MiB, and
every bfloat16 operator is refused.

`DeviceName.cpp` exports `cnDeviceGetName` ahead of the driver.  A name CNNL
recognizes (`MLU` followed by a digit) passes through; any other is replaced
by the series name for the chip's compute capability (`MLU590` for 5.x,
`MLU370` for 3.x).  Set `TP_MLU_DEVICE_NAME` to force a specific name.  The
replacement only takes effect when this library is loaded before anything
else that links the driver (such as another framework's MLU integration) in
the same process.

## Profiling

```python
from tensorplay.profiler import profile, ProfilerActivity

with profile(activities=[ProfilerActivity.CPU, ProfilerActivity.PrivateUse1]) as prof:
    train_step()
print(prof.key_averages().table(sort_by="device_time_total", row_limit=10))
```

Each operation gets its MLU time (notifiers on the device's queue) and, when
`libcnpapi.so` loads, the kernels, copies and fills it launched
(`prof.kineto_results()`, exported traces).  `profile(use_device="mlu")` asks
for the same.  `tensorplay.utils.count_host_fallbacks()` lists the operators
that still run on the host.

## Repository layout

| Path | Role |
| --- | --- |
| `tensorplay_mlu/csrc/MLURuntime.*` | errors, current device, one queue per device, per-thread CNNL handles |
| `tensorplay_mlu/csrc/MLUAllocator.*` | caching allocator over `cnrtMalloc`, memory statistics |
| `tensorplay_mlu/csrc/MLUHooks.cpp` | device count, synchronization, byte copies, memory info |
| `tensorplay_mlu/csrc/Cnnl.*` | tensor descriptors, workspaces, dense/strided copies, dtype rules |
| `tensorplay_mlu/csrc/DeviceName.cpp` | the product name the kernel library selects its kernels by |
| `tensorplay_mlu/csrc/CopyKernels.cpp` | `empty`, `*_like` factories, `copy_`, `fill_`, `clone` |
| `tensorplay_mlu/csrc/BinaryKernels.cpp` | add/sub/mul/div with broadcasting, in-place and scalar forms |
| `tensorplay_mlu/csrc/UnaryKernels.cpp` | exp/log/sqrt/..., `pow`, activations and their backward kernels |
| `tensorplay_mlu/csrc/BlasKernels.cpp` | `mm`, `addmm`, `bmm`, `matmul`, `linear` |
| `tensorplay_mlu/csrc/ReduceKernels.cpp` | `sum`, `mean`, `amax`/`amin`, softmax/log-softmax with backward |
| `tensorplay_mlu/csrc/ConvKernels.cpp` | convolutions (1D/2D/3D, transposed, dilated, grouped) with backward |
| `tensorplay_mlu/csrc/PoolKernels.cpp` | max/average pooling (1D/2D/3D) and adaptive variants with backward |
| `tensorplay_mlu/csrc/BatchNormKernels.cpp` | batch normalization in training and evaluation |
| `tensorplay_mlu/csrc/IndexKernels.cpp` | indexing, scatter, gather, scan and `where` |
| `tensorplay_mlu/csrc/Cncl.cpp` | CNCL collective entry points over MLU device pointers |
| `tensorplay_mlu/distributed.py` | the `cncl` process group that drives them |
| `tensorplay_mlu/csrc/MLUProfiler.cpp` | profiler device time: a notifier pair around each operation |
| `tensorplay_mlu/csrc/MLUActivity.cpp` | profiler kernel-level records through CNPAPI |
| `tensorplay_mlu/csrc/Module.cpp` | C entry points: host fallbacks, memory statistics |
| `tensorplay_mlu/csrc/NNKernels.cpp` | neural-network kernels: layer norm, dropout, embedding, NLL loss (forward/backward) |
| `tensorplay_mlu/compiler/` | MagicMind graph backend: ONNX lowering, content-addressed model cache, MLU execution |
| `tests/` | kernel-by-kernel comparison against the host |

## Versioning and compatibility

`tensorplay-mlu` is versioned independently of TensorPlay (semver, see
`pyproject.toml`).  It targets the out-of-tree device slot TensorPlay
reserves for accelerators, so the two projects move at their own pace:

| Component | Version / branch |
| --- | --- |
| `tensorplay-mlu` | `0.1.0` (this repository, `main`) |
| TensorPlay | any release that ships the `PrivateUse1` slot (`pip install tensorplay`) |
| Neuware toolkit | `CNToolkit` + `CNNL` from the Cambricon developer site |
| Tested card | MLU590 series (virtualized, generic product name) |

The native sources are compiled against the installed toolkit on first load
and cached under `~/.cache/tensorplay_extensions`, so a `pip install -U
tensorplay-mlu` picks up new kernels without rebuilding TensorPlay.

## Development

The test suite compares every kernel against the host and needs a Cambricon
MLU with the Neuware toolkit:

```bash
pip install pytest
pytest tests
```

How the out-of-tree slot works is documented in [Adding a device backend] in
the TensorPlay repository.

## License

[Apache License 2.0](LICENSE).

*Cambricon and the Cambricon logo are trademarks of Cambricon Technologies
Corporation Limited.  This project is not affiliated with or endorsed by
Cambricon.*

[TensorPlay]: https://github.com/lexing-2026/TensorPlay
[Adding a device backend]: https://github.com/lexing-2026/TensorPlay/blob/main/docs/source/notes/backends.md
