Issue Station
Sign In
Tools hereToolbox

pytorch - 💡(How to fix) Fix torch.cuda.nccl.broadcast silently ignores root argument — always broadcasts from index 0 [1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
pytorch/pytorch#179908•Fetched 2026-04-11 06:11:55
View on GitHub
Comments
0
Participants
1
Timeline
26
Reactions
2
Author
lodestone-rocklodestone-rock
Participants
lodestone-rocklodestone-rock
Timeline (top)
mentioned ×12subscribed ×12labeled ×2

Root Cause

Root cause: In torch/csrc/cuda/nccl.cpp, the root rank is hardcoded as 0 in the ncclBcast call instead of using the root parameter.

Fix Action

Fix / Workaround

Workaround: Swap tensors[root] to index 0 before calling with root=0, then swap back.

CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 43 bits physical, 48 bits virtual Byte Order: Little Endian CPU(s): 128 On-line CPU(s) list: 0-127 Vendor ID: AuthenticAMD Model name: AMD EPYC 7H12 64-Core Processor CPU family: 23 Model: 49 Thread(s) per core: 2 Core(s) per socket: 64 Socket(s): 1 Stepping: 0 Frequency boost: enabled CPU(s) scaling MHz: 62% CPU max MHz: 2600.0000 CPU min MHz: 1500.0000 BogoMIPS: 5190.13 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sev sev_es ibpb_exit_to_user Virtualization: AMD-V L1d cache: 2 MiB (64 instances) L1i cache: 2 MiB (64 instances) L2 cache: 32 MiB (64 instances) L3 cache: 256 MiB (16 instances) NUMA node(s): 1 NUMA node0 CPU(s): 0-127 Vulnerability Gather data sampling: Not affected Vulnerability Indirect target selection: Not affected Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Not affected Vulnerability Reg file data sampling: Not affected Vulnerability Retbleed: Mitigation; untrained return thunk; SMT enabled with STIBP protection Vulnerability Spec rstack overflow: Mitigation; Safe RET Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Retpolines; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected Vulnerability Srbds: Not affected Vulnerability Tsa: Not affected Vulnerability Tsx async abort: Not affected Vulnerability Vmscape: Mitigation; IBPB before exit to userspace

Code Example

import torch
import torch.cuda.nccl as nccl

tensors = [torch.ones(1).cuda(i) * float(i * 10) for i in range(torch.cuda.device_count())]
nccl.broadcast(tensors, root=1)

# Expected: [tensor([10.], device='cuda:0'), tensor([10.], device='cuda:1')]
# Actual: [tensor([0.], device='cuda:0'), tensor([0.], device='cuda:1')]

---

PyTorch version: 2.9.0+cu128
Is debug build: False
CUDA used to build PyTorch: 12.8
ROCM used to build PyTorch: N/A

OS: Ubuntu 24.04.4 LTS (x86_64)
GCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0
Clang version: Could not collect
CMake version: Could not collect
Libc version: glibc-2.39

Python version: 3.12.3 (main, Mar  3 2026, 12:15:18) [GCC 13.3.0] (64-bit runtime)
Python platform: Linux-6.8.0-106-generic-x86_64-with-glibc2.39
Is CUDA available: True
CUDA runtime version: 12.0.140
CUDA_MODULE_LOADING set to: 
GPU models and configuration: 
GPU 0: NVIDIA RTX PRO 6000 Blackwell Server Edition
GPU 1: NVIDIA RTX PRO 6000 Blackwell Server Edition

Nvidia driver version: 590.44.01
cuDNN version: Could not collect
Is XPU available: False
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
Caching allocator config: N/A

CPU:
Architecture:                            x86_64
CPU op-mode(s):                          32-bit, 64-bit
Address sizes:                           43 bits physical, 48 bits virtual
Byte Order:                              Little Endian
CPU(s):                                  128
On-line CPU(s) list:                     0-127
Vendor ID:                               AuthenticAMD
Model name:                              AMD EPYC 7H12 64-Core Processor
CPU family:                              23
Model:                                   49
Thread(s) per core:                      2
Core(s) per socket:                      64
Socket(s):                               1
Stepping:                                0
Frequency boost:                         enabled
CPU(s) scaling MHz:                      62%
CPU max MHz:                             2600.0000
CPU min MHz:                             1500.0000
BogoMIPS:                                5190.13
Flags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sev sev_es ibpb_exit_to_user
Virtualization:                          AMD-V
L1d cache:                               2 MiB (64 instances)
L1i cache:                               2 MiB (64 instances)
L2 cache:                                32 MiB (64 instances)
L3 cache:                                256 MiB (16 instances)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-127
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Mitigation; untrained return thunk; SMT enabled with STIBP protection
Vulnerability Spec rstack overflow:      Mitigation; Safe RET
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; Retpolines; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Mitigation; IBPB before exit to userspace

Versions of relevant libraries:
[pip3] mypy_extensions==1.1.0
[pip3] numpy==2.4.4
[pip3] nvidia-cublas==13.1.0.3
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti==13.0.85
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc==13.0.88
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime==13.0.96
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-cu13==9.19.0.56
[pip3] nvidia-cufft==12.0.0.61
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-curand==10.4.0.35
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver==12.0.4.66
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse==12.6.3.3
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cusparselt-cu13==0.8.0
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nccl-cu13==2.28.9
[pip3] nvidia-nvjitlink==13.0.88
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvtx==13.0.85
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] torch==2.9.0
[pip3] torchvision==0.24.0
[pip3] triton==3.5.0
RAW_BUFFERClick to expand / collapse

🐛 Describe the bug

nccl.broadcast(tensors, root=N) silently ignores root and always broadcasts from tensors[0]. nccl.reduce with root=N works correctly, making this inconsistent.

Root cause: In torch/csrc/cuda/nccl.cpp, the root rank is hardcoded as 0 in the ncclBcast call instead of using the root parameter.

Minimal reproducer:

import torch
import torch.cuda.nccl as nccl

tensors = [torch.ones(1).cuda(i) * float(i * 10) for i in range(torch.cuda.device_count())]
nccl.broadcast(tensors, root=1)

# Expected: [tensor([10.], device='cuda:0'), tensor([10.], device='cuda:1')]
# Actual: [tensor([0.], device='cuda:0'), tensor([0.], device='cuda:1')]

Workaround: Swap tensors[root] to index 0 before calling with root=0, then swap back.

Versions:

PyTorch version: 2.9.0+cu128
Is debug build: False
CUDA used to build PyTorch: 12.8
ROCM used to build PyTorch: N/A

OS: Ubuntu 24.04.4 LTS (x86_64)
GCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0
Clang version: Could not collect
CMake version: Could not collect
Libc version: glibc-2.39

Python version: 3.12.3 (main, Mar  3 2026, 12:15:18) [GCC 13.3.0] (64-bit runtime)
Python platform: Linux-6.8.0-106-generic-x86_64-with-glibc2.39
Is CUDA available: True
CUDA runtime version: 12.0.140
CUDA_MODULE_LOADING set to: 
GPU models and configuration: 
GPU 0: NVIDIA RTX PRO 6000 Blackwell Server Edition
GPU 1: NVIDIA RTX PRO 6000 Blackwell Server Edition

Nvidia driver version: 590.44.01
cuDNN version: Could not collect
Is XPU available: False
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
Caching allocator config: N/A

CPU:
Architecture:                            x86_64
CPU op-mode(s):                          32-bit, 64-bit
Address sizes:                           43 bits physical, 48 bits virtual
Byte Order:                              Little Endian
CPU(s):                                  128
On-line CPU(s) list:                     0-127
Vendor ID:                               AuthenticAMD
Model name:                              AMD EPYC 7H12 64-Core Processor
CPU family:                              23
Model:                                   49
Thread(s) per core:                      2
Core(s) per socket:                      64
Socket(s):                               1
Stepping:                                0
Frequency boost:                         enabled
CPU(s) scaling MHz:                      62%
CPU max MHz:                             2600.0000
CPU min MHz:                             1500.0000
BogoMIPS:                                5190.13
Flags:                                   fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sev sev_es ibpb_exit_to_user
Virtualization:                          AMD-V
L1d cache:                               2 MiB (64 instances)
L1i cache:                               2 MiB (64 instances)
L2 cache:                                32 MiB (64 instances)
L3 cache:                                256 MiB (16 instances)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-127
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Mitigation; untrained return thunk; SMT enabled with STIBP protection
Vulnerability Spec rstack overflow:      Mitigation; Safe RET
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; Retpolines; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Mitigation; IBPB before exit to userspace

Versions of relevant libraries:
[pip3] mypy_extensions==1.1.0
[pip3] numpy==2.4.4
[pip3] nvidia-cublas==13.1.0.3
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti==13.0.85
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc==13.0.88
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime==13.0.96
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-cu13==9.19.0.56
[pip3] nvidia-cufft==12.0.0.61
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-curand==10.4.0.35
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver==12.0.4.66
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse==12.6.3.3
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cusparselt-cu13==0.8.0
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nccl-cu13==2.28.9
[pip3] nvidia-nvjitlink==13.0.88
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvtx==13.0.85
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] torch==2.9.0
[pip3] torchvision==0.24.0
[pip3] triton==3.5.0

cc @awgu @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @pragupta @msaroufim @dcci @aditvenk @xmfan

extent analysis

TL;DR

The nccl.broadcast function can be fixed by swapping tensors[root] to index 0 before calling with root=0, then swapping back, as a temporary workaround.

Guidance

  • The root cause of the issue is the hardcoded root rank as 0 in the ncclBcast call, which can be verified by checking the torch/csrc/cuda/nccl.cpp file.
  • To mitigate the issue, swap tensors[root] to index 0 before calling nccl.broadcast with root=0, then swap back after the call.
  • The provided minimal reproducer can be used to test the workaround.
  • It is recommended to update the torch/csrc/cuda/nccl.cpp file to use the root parameter instead of hardcoding it to 0.

Example

import torch
import torch.cuda.nccl as nccl

tensors = [torch.ones(1).cuda(i) * float(i * 10) for i in range(torch.cuda.device_count())]
root = 1
tensors[0], tensors[root] = tensors[root], tensors[0]  # swap
nccl.broadcast(tensors, root=0)
tensors[0], tensors[root] = tensors[root], tensors[0]  # swap back

Notes

The provided workaround is a temporary solution and may not be suitable for all use cases. It is recommended to update the torch/csrc/cuda/nccl.cpp file to fix the issue permanently.

Recommendation

Apply the workaround by swapping tensors[root] to index 0 before calling nccl.broadcast with root=0, then swapping back, as it provides a temporary solution to the issue.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#permission error#memory optimization#batch processing#GPU compatibility

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING

pytorch - 💡(How to fix) Fix torch.cuda.nccl.broadcast silently ignores root argument — always broadcasts from index 0 [1 participants]