Skip to content

chore(qy): pin InfiniRT/Ops/CCL for QY MoE bf16-warpnk (2026.11.12) #1583

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Ceng23333 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

chore(qy): pin InfiniRT/Ops/CCL for QY MoE bf16-warpnk (2026.11.12) #1583

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion submodules/InfiniCCL
Open in desktop
Submodule InfiniCCL updated 75 files
+4 −0 .github/pull_request_template.md
+1 −2 .github/workflows/clang-format.yml
+94 −544 CMakeLists.txt
+7 −2 README.md
+40 −0 cmake/backends/CNCL.cmake
+70 −0 cmake/backends/HCCL.cmake
+63 −0 cmake/backends/MCCL.cmake
+29 −0 cmake/backends/MPI.cmake
+92 −0 cmake/backends/NCCL.cmake
+56 −0 cmake/devices/Ascend.cmake
+53 −0 cmake/devices/Cambricon.cmake
+28 −0 cmake/devices/DeviceDetection.cmake
+98 −0 cmake/devices/Hygon.cmake
+54 −0 cmake/devices/Iluvatar.cmake
+51 −0 cmake/devices/MetaX.cmake
+89 −0 cmake/devices/Moore.cmake
+25 −0 cmake/devices/Nvidia.cmake
+46 −0 cmake/devices/Qy.cmake
+85 −0 cmake/devices/THead.cmake
+9 −0 examples/CMakeLists.txt
+221 −0 examples/ccl/all_to_all.cc
+205 −0 examples/ccl/gather.cc
+196 −0 examples/ccl/reduce.cc
+1 −0 examples/ccl/reduce_scatter.cc
+199 −0 examples/ccl_mpi_hybrid/all_to_all.cc
+188 −0 examples/ccl_mpi_hybrid/gather.cc
+172 −0 examples/ccl_mpi_hybrid/reduce.cc
+1 −1 include/comm.h
+3 −0 scripts/gen_bridge.py
+39 −1 src/CMakeLists.txt
+5 −0 src/backend.h
+8 −0 src/backend_device_map.h
+99 −0 src/backends/ccl/common/impl/all_to_all.h
+67 −0 src/backends/ccl/common/impl/gather.h
+57 −0 src/backends/ccl/common/impl/reduce.h
+193 −0 src/backends/ccl/hccl/api.h
+15 −0 src/backends/ccl/hccl/ascend/api.h
+268 −0 src/backends/ccl/hccl/ascend/average.cc
+24 −0 src/backends/ccl/hccl/ascend/average.h
+36 −0 src/backends/ccl/hccl/checks.h
+17 −0 src/backends/ccl/hccl/impl/all_gather.h
+17 −0 src/backends/ccl/hccl/impl/all_reduce.h
+17 −0 src/backends/ccl/hccl/impl/comm_destroy.h
+17 −0 src/backends/ccl/hccl/impl/comm_init_rank.h
+17 −0 src/backends/ccl/hccl/impl/get_unique_id.h
+17 −0 src/backends/ccl/hccl/impl/recv.h
+17 −0 src/backends/ccl/hccl/impl/reduce_scatter.h
+17 −0 src/backends/ccl/hccl/impl/send.h
+16 −0 src/backends/ccl/hccl/reduction_op.h
+70 −0 src/backends/ccl/hccl/type_map.h
+41 −0 src/backends/ccl/mccl/api.h
+17 −0 src/backends/ccl/mccl/impl/all_gather.h
+17 −0 src/backends/ccl/mccl/impl/all_to_all.h
+17 −0 src/backends/ccl/mccl/impl/gather.h
+17 −0 src/backends/ccl/mccl/impl/recv.h
+17 −0 src/backends/ccl/mccl/impl/reduce.h
+17 −0 src/backends/ccl/mccl/impl/reduce_scatter.h
+17 −0 src/backends/ccl/mccl/impl/send.h
+17 −0 src/backends/ccl/nccl/api.h
+17 −0 src/backends/ccl/nccl/impl/all_to_all.h
+17 −0 src/backends/ccl/nccl/impl/gather.h
+17 −0 src/backends/ccl/nccl/impl/reduce.h
+15 −0 src/backends/ccl/nccl/qy/api.h
+4 −1 src/base/all_to_all.h
+10 −10 src/base/gather.h
+10 −9 src/base/reduce.h
+10 −0 src/device.h
+31 −0 src/devices/ascend/caster_.h
+43 −0 src/devices/ascend/checks.h
+24 −0 src/devices/ascend/data_type_.h
+31 −0 src/devices/ascend/device_.h
+84 −0 src/devices/ascend/runtime_.h
+26 −0 src/devices/qy/data_type_.h
+32 −0 src/devices/qy/device_.h
+75 −0 src/devices/qy/runtime_.h
2 changes: 1 addition & 1 deletion submodules/InfiniOps
Open in desktop
Submodule InfiniOps updated 67 files
+94 −9 CMakeLists.txt
+2 −0 cmake/InfiniOpsConfig.cmake.in
+15 −7 docs/build-nvidia-inference.md
+26 −0 include/moe_align_kernel.h
+16 −0 include/moe_sum_kernel.h
+22 −0 include/moe_topk_sigmoid_kernel.h
+16 −0 include/mul_scalar_kernel.h
+20 −0 include/rope_kernel.h
+22 −0 include/rope_pack_kernel.h
+16 −0 include/silu_and_mul_kernel.h
+11 −0 scripts/dev/platforms.sh
+78 −0 scripts/dlcc_wrapper.sh
+33 −3 src/CMakeLists.txt
+1 −1 src/device.h
+2 −0 src/linked/torch/qy/flash_attn.yaml
+44 −0 src/linked/torch/qy/ops/flash_attn_varlen_func/flash_attn.cc
+49 −0 src/linked/torch/qy/ops/flash_attn_varlen_func/flash_attn.h
+4 −0 src/linked/torch/qy/ops/flash_attn_varlen_func/flash_attn.yaml
+50 −0 src/linked/torch/qy/ops/flash_attn_with_kvcache/flash_attn.cc
+50 −0 src/linked/torch/qy/ops/flash_attn_with_kvcache/flash_attn.h
+4 −0 src/linked/torch/qy/ops/flash_attn_with_kvcache/flash_attn.yaml
+36 −0 src/native/cuda/qy/blas.h
+26 −0 src/native/cuda/qy/blas_utils.h
+77 −0 src/native/cuda/qy/caster.cuh
+13 −0 src/native/cuda/qy/data_type_.h
+6 −0 src/native/cuda/qy/device_.h
+42 −0 src/native/cuda/qy/device_property.h
+212 −0 src/native/cuda/qy/moe_align_kernel.cu
+26 −0 src/native/cuda/qy/moe_align_kernel.h
+149 −0 src/native/cuda/qy/moe_sum_kernel.cu
+16 −0 src/native/cuda/qy/moe_sum_kernel.h
+317 −0 src/native/cuda/qy/moe_topk_sigmoid_kernel.cu
+22 −0 src/native/cuda/qy/moe_topk_sigmoid_kernel.h
+90 −0 src/native/cuda/qy/mul_scalar_kernel.cu
+16 −0 src/native/cuda/qy/mul_scalar_kernel.h
+19 −0 src/native/cuda/qy/ops/add/kernel.h
+19 −0 src/native/cuda/qy/ops/argmax/kernel.h
+21 −0 src/native/cuda/qy/ops/convolution/kernel.h
+19 −0 src/native/cuda/qy/ops/copy/kernel.h
+19 −0 src/native/cuda/qy/ops/embedding/kernel.h
+21 −0 src/native/cuda/qy/ops/fill/kernel.h
+19 −0 src/native/cuda/qy/ops/fused_add_rms_norm/kernel.h
+21 −0 src/native/cuda/qy/ops/gelu/kernel.h
+18 −0 src/native/cuda/qy/ops/gemm/cublas.h
+21 −0 src/native/cuda/qy/ops/mul/kernel.h
+21 −0 src/native/cuda/qy/ops/relu/kernel.h
+20 −0 src/native/cuda/qy/ops/reshape_and_cache_flash/kernel.h
+19 −0 src/native/cuda/qy/ops/rms_norm/kernel.h
+19 −0 src/native/cuda/qy/ops/rotary_embedding/kernel.h
+21 −0 src/native/cuda/qy/ops/sigmoid/kernel.h
+21 −0 src/native/cuda/qy/ops/silu/kernel.h
+19 −0 src/native/cuda/qy/ops/silu_and_mul/kernel.h
+21 −0 src/native/cuda/qy/ops/softmax/kernel.h
+21 −0 src/native/cuda/qy/ops/topk_softmax/kernel.h
+172 −0 src/native/cuda/qy/rope_kernel.cu
+20 −0 src/native/cuda/qy/rope_kernel.h
+148 −0 src/native/cuda/qy/rope_pack_kernel.cu
+19 −0 src/native/cuda/qy/rope_pack_kernel.h
+9 −0 src/native/cuda/qy/runtime_.h
+15 −0 src/native/cuda/qy/runtime_utils.h
+114 −0 src/native/cuda/qy/silu_and_mul_kernel.cu
+16 −0 src/native/cuda/qy/silu_and_mul_kernel.h
+1 −0 src/pybind11_utils.h
+5 −0 src/torch/device_.h
+10 −1 src/torch/instantiate.h
+28 −0 src/torch/qy/c10.h
+1 −0 tests/conftest.py
2 changes: 1 addition & 1 deletion submodules/InfiniRT
Open in desktop
Submodule InfiniRT updated 15 files
+54 −4 CMakeLists.txt
+27 −0 cmake/InfiniRTConfig.cmake.in
+78 −0 scripts/dlcc_wrapper.sh
+7 −0 scripts/generate_public_headers.py
+1 −0 scripts/run_performance_tests.py
+7 −0 src/CMakeLists.txt
+4 −1 src/device.h
+30 −0 src/native/cuda/qy/data_type_.h
+13 −0 src/native/cuda/qy/device_.h
+188 −0 src/native/cuda/qy/runtime_.h
+18 −0 tests/CMakeLists.txt
+1 −1 tests/install_consumer/CMakeLists.txt
+5 −1 tests/install_consumer_smoke.cc
+2 −0 tests/performance/CMakeLists.txt
+5 −0 tests/test_runtime_dispatch.cc