feat: support CCL Reduce

[GordonYang1]

Summary

This PR adds Reduce support to the shared CCL backend abstraction, including native bindings through the existing NCCL and MCCL provider layers for the existing public infinicclReduce() API. It also keeps inter-only communicators on the existing OpenMPI staging path during mixed-backend bootstrap flows, hardens the shared OpenMPI/MPICH Reduce implementation with typed-reduction validation, overflow checks, automatic host-buffer cleanup, and correct integer Avg scaling, defines zero-count behavior, makes the existing MPI example propagate validation failures, and includes CCL-only plus OpenMPI-assisted Reduce examples with correctness and communication-bandwidth reporting.

Changes

• Public API and Dispatch

  • Enable the existing infinicclReduce() API for configured CCL backends through generated bridge dispatch without changing its public signature.
  • Use a matching native CCL intra communicator when available, and otherwise delegate to the existing OpenMPI provider when the communicator has only an OpenMPI inter communicator.
  • Return success for a zero-count Reduce after validating the communicator, data type, reduction operation, and root, without requiring non-null buffers or entering a backend.

• Common CCL Implementation

  • Add a shared provider-oriented CCL Reduce implementation following the existing AllReduce structure.
  • Validate the native communicator backend, device, rank, world size, root, and handle before dispatch.
  • Map InfiniCCL data types and reduction operations through the configured provider and return NotSupported when the provider cannot represent either value.

• Existing CCL Provider Bindings

  • Extend the existing NCCL and MCCL API wrappers with thin bindings to ncclReduce() and mcclReduce().
  • Register Reduce with the existing NCCL and MCCL provider layers.

• MPI Correctness and Safety

  • Keep arithmetic reduction on typed MPI_Reduce() calls and reject data types that map to MPI_BYTE, avoiding invalid byte-wise reductions for Float16 and BFloat16.
  • Add communicator, world-size, size_t byte-size, and MPI int count-range validation.
  • Manage host staging buffers with automatic cleanup across success and error paths.
  • Scale integer Avg results in floating point before converting them back to the destination type.

• Examples, Validation, and Metrics

  • Add a thread-per-GPU single-node CCL Reduce example using one shared unique ID and rank-based communicator initialization.
  • Add an OpenMPI-assisted CCL Reduce example that first exercises the MPI fallback through an inter-only communicator, then initializes the native CCL communicator for GPU reduction.
  • Update the existing MPI example to validate the complete root result and propagate failure through the process exit status.
  • Report per-iteration time, algorithm bandwidth, and bus bandwidth using the nccl-tests Reduce convention, whose bus correction factor is 1.
  • Parse numeric options without exceptions and guard example buffer-size calculations against overflow.

Platform and Backend Affected

Platform

• CPU
• NVIDIA GPU
• Iluvatar GPU
• MetaX GPU
• Moore Threads GPU
• Cambricon MLU
• HYGON DCU

Backend

• OpenMPI
• MPICH
• NCCL
• MCCL

Performance Impact

• No performance impact
• Performance improved
• Performance regression possible

This adds GPU-native NCCL and MCCL paths for Reduce, avoiding the existing MPI host-staging path when a supported CCL backend and matching native communicator are available. The shared MPI path remains host-staged and retains typed reduction semantics with additional validation and cleanup. Other collective operations are intended to remain unchanged. The validation-log timings below are execution references rather than a quantified cross-backend performance comparison.

Known Issues & Future Work

• The CCL collective backend on this branch adds Reduce alongside the existing AllReduce; other independently developed CCL collectives are outside this branch.
• Reduce inherits the backend/device combinations, data types, and reduction operations supported by the existing CCL providers; this PR does not add a new provider or device integration.
• The server examples validate Float32 Sum payloads on the default stream. Additional data types, reduction operations, non-default streams, in-place layouts, and runtime coverage on Iluvatar and Moore Threads remain future work.
• The shared OpenMPI/MPICH fallback remains host-staged with per-call allocations, rejects Float16 and BFloat16 arithmetic reductions until a typed host-conversion path is available, and rejects counts above the MPI int range; chunked transfers remain future work.

Test Results

The implementation commit is 5d8ac19c11e989caded690b75ed884653a25bfa5, a single commit directly based on ef4045a2d99837c75c2acd90aae57dacb83172d2. The attached final evidence contains exactly 15 hash-verified canonical logs: all 15 targets passed, with Correct: YES 17 times and Correct: NO 0 times. The extra two positive results are the additional expected validation cases in the broadcast log.

All four current-commit Reduce paths passed with a 1,048,576-element Float32 Sum payload (4.00 MiB), 2 warm-up iterations, and 20 profiled iterations:

Path	Test topology	Expect / Actual	Time	Alg BW	Bus BW
Pure NCCL	8 NVIDIA A100 GPUs	36 / 36	0.044 ms	94.34 GB/s	94.34 GB/s
OpenMPI + NCCL hybrid	8 NVIDIA A100 GPUs	36 / 36	0.047 ms	89.56 GB/s	89.56 GB/s
Heterogeneous OpenMPI	8 NVIDIA A100 + 8 MetaX C550 GPUs	136 / 136	13.441 ms	0.31 GB/s	0.31 GB/s
Pure MCCL	8 MetaX C550 GPUs	36 / 36	0.438 ms	9.57 GB/s	9.57 GB/s

The reported times are root-rank averages over the 20 profiled calls after the 2 warm-up calls. Reduce uses a bus-bandwidth correction factor of 1, so the reported algorithm and bus bandwidths are equal. The values were independently recomputed while accounting for the displayed time being rounded to three decimal places. They are included as execution evidence, not as a cross-platform benchmark comparison.

• The pure NCCL and OpenMPI+NCCL configurations each passed both selected targets, 2/2 and 2/2, on all 8 A100 GPUs.
• The heterogeneous OpenMPI configuration passed all 9 MPI targets on 16 ranks. Ranks 0-7 mapped to NVIDIA devices 0-7, and ranks 8-15 mapped to MetaX devices 0-7.
• The pure MCCL configuration passed both selected targets on all 8 MetaX C550 GPUs, with ranks 0-7 mapped to devices 0-7.
• All five formal invocations returned zero. There were no harness retries, MetaX vendor queue retries, timeouts, missing canonical logs, or recorded finalization errors.

Test Involved Platform

• CPU
• NVIDIA GPU
• Iluvatar GPU
• MetaX GPU
• Moore Threads GPU
• Cambricon MLU
• HYGON DCU

Test Involved Backend

• OpenMPI
• MPICH
• NCCL
• MCCL

Pure CCL (NCCL) on single-node NVIDIA:
ccl_all_reduce.log
ccl_reduce.log

CCL + MPI on single-node NVIDIA:
ccl_mpi_hybrid_all_reduce.log
ccl_mpi_hybrid_reduce.log

MPI on Heterogeneous Cluster:
mpi_all_gather.log
mpi_all_reduce.log
mpi_all_to_all.log
mpi_broadcast.log
mpi_gather.log
mpi_reduce.log
mpi_reduce_scatter.log
mpi_scatter.log
mpi_send_recv.log

Pure CCL (MCCL) on single-node MetaX:
ccl_all_reduce.log
ccl_reduce.log

---

Checklist

Every contributor must verify every item below before requesting
review. Tick each box only after the check has actually been performed —
do not tick speculatively. If an item truly does not apply, replace the
checkbox with N/A and briefly explain why in an inline comment.

Title, Branch, and Commits

• PR title follows Conventional Commits (e.g. feat: …, fix(nccl): …).
• Branch name follows <type>/xxx-yyyy-zzzz where <type> matches the PR title's Conventional Commits type and words are joined with hyphens (see CONTRIBUTING.md §Branches).
• Each commit message follows Conventional Commits.
• Small PR is a single squashable commit; or, for a large PR, every commit is meaningful, well-formed, and independently reviewable (see CONTRIBUTING.md §Pull Requests).
• No stray merge commits from master — the branch is rebased cleanly on top of the current master.
• No fixup! / squash! / wip commits remain.

Scope and Design

• Changes are minimal — no unrelated modifications were introduced (CONTRIBUTING.md §Code/General).
• No dead code, commented-out blocks, debug prints, printf/std::cout/print(...) left behind, or TODO without an owner and issue link.
• No unrelated formatting churn that would obscure the diff.
• Public API changes (if any) are intentional, documented, and reflected in affected callers/tests.

General Code Hygiene

• The code is self-explanatory; comments were added only where the intent or rationale is non-obvious (CONTRIBUTING.md §Code/General).
• Every modified or added file ends with a single trailing newline (CONTRIBUTING.md §Code/General).
• No trailing whitespace, inconsistent indentation, or mixed formatting styles remain.
• Identifiers referenced in comments or error messages are wrapped in Markdown backticks (e.g. the `AllReduce` implementation) (CONTRIBUTING.md §Code/General).
• All comments and error messages are in English (CONTRIBUTING.md §Code/General).
• Comments and error messages are complete sentences — capitalized first letter, terminal punctuation — unless the language/framework convention says otherwise (CONTRIBUTING.md §Code/General; §Python).

C++ Specific (if C++ files changed)

• Code follows the Google C++ Style Guide strictly.
• clang-format (version 16, per .github/workflows/clang-format.yml) has been run against all modified applicable files; the diff is clean.
• No exceptions are thrown. Error paths use assert with messages that include at least __FILE__, __LINE__, and __func__ (CONTRIBUTING.md §C++).
• Error and warning message wording follows the LLVM Coding Standards (CONTRIBUTING.md §C++).
• N/A- Constructor initializer list order matches member declaration order (CONTRIBUTING.md §C++).
• Exactly one blank line between classes, between classes and functions, and between functions (CONTRIBUTING.md §C++).
• Exactly one blank line between members (functions and variables) within a class (CONTRIBUTING.md §C++).
• Exactly one blank line before and after the contents of a namespace (CONTRIBUTING.md §C++).

Python Specific (if Python files changed)

• N/A- Code is PEP 8 compliant; ruff check passes cleanly on CI (see `.github/workflows/ruff.yml).
• N/A- ruff format --check passes cleanly — if not, run ruff format and commit the result.
• N/A- Comments are complete English sentences, starting with a capital letter and ending with punctuation; Markdown backticks are used for code references (CONTRIBUTING.md §Python).
• N/A- Framework-specific conventions (e.g. lowercase pytest.skip messages without terminal period) are honored where applicable (CONTRIBUTING.md §Python).
• N/A- No blank line between the function signature and the body when there is no docstring or comment (CONTRIBUTING.md §Python).
• N/A- A blank line is present before and after if, for, and similar control-flow statements (CONTRIBUTING.md §Python).
• N/A- A blank line appears before each return, except when it directly follows a control-flow statement (CONTRIBUTING.md §Python).
• N/A- Docstrings (if any) follow PEP 257 conventions.
• N/A- Type hints are added / kept consistent with the surrounding code.

Testing

• All applicable example programs have been built and tested successfully on at least one supported heterogeneous cluster setup.

Build, CI, and Tooling

• N/A- New backends or devices have been added to auto-detection in CMakeLists.txt under if(AUTO_DETECT_DEVICES) or to if(AUTO_DETECT_BACKENDS) if applicable.
• Both CI workflows (clang-format.yml, ruff.yml) are green locally (or expected to be green on CI).

Documentation

• N/A- README.md, CONTRIBUTING.md, or inline docs updated when behavior, build flags, or developer workflow changed.
• N/A- Any user-visible breaking change is called out explicitly under "Summary" and in the commit/PR title with a ! or BREAKING CHANGE: footer.

Security and Safety

• No secrets, access tokens, internal URLs, customer data, or personal hardware identifiers have been committed.
• N/A- Third-party code is license-compatible and attributed.
• No unsafe pointer arithmetic, uninitialized reads, or missing bounds checks were introduced.