I’ve been having an issue doing llama 8b pre-training FSDP 2 with an on-prem single H200x8 bare metal instance, where I’m getting very jittery performance from inexplicably slow cpu ops that take a couple seconds before enqueuing any CUDA kernels. I’ve profiled an example of a single rank, where you can see it do be the case for aten::chunk_cat where it takes 2.5 seconds, while other instances of the aten::chunk_cat in other iterations only take like 2ms. The next highest was only 250ms. I’m really confused since it looks like none of the gpu streams are busy during this time either.
From many other profiles i’ve taken and sampled at different iterations, it doesn’t seem to be specific to this operation. I’ve ruled out any sort of data starvation issues, since this always happen in the middle of the fwd/bwd passes and my WandB metrics from torchtitan are showing only 1% time spent on dataloading. any suggestions or other things I should check would be very helpful! from what i could tell from also profiling via nsight, i couldn’t find any sort of thread synchronization issues that might also cause this.
(titan) root@se-h1-17-gpu:~/torchtitan# python -m torch.utils.collect_env
/root/miniconda3/envs/titan/lib/python3.10/runpy.py:126: RuntimeWarning: 'torch.utils.collect_env' found in sys.modules after import of package 'torch.utils', but prior to execution of 'torch.utils.collect_env'; this may result in unpredictable behaviour
warn(RuntimeWarning(msg))
Collecting environment information...
PyTorch version: 2.10.0.dev20250926+cu128
Is debug build: False
CUDA used to build PyTorch: 12.8
ROCM used to build PyTorch: N/A
OS: Ubuntu 22.04.5 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04.2) 11.4.0
Clang version: Could not collect
CMake version: version 3.22.1
Libc version: glibc-2.35
Python version: 3.10.18 (main, Jun 5 2025, 13:14:17) [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-5.15.0-151-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.9.86
CUDA_MODULE_LOADING set to:
GPU models and configuration:
GPU 0: NVIDIA H200
GPU 1: NVIDIA H200
GPU 2: NVIDIA H200
GPU 3: NVIDIA H200
GPU 4: NVIDIA H200
GPU 5: NVIDIA H200
GPU 6: NVIDIA H200
GPU 7: NVIDIA H200
Nvidia driver version: 575.57.08
cuDNN version: Could not collect
Is XPU available: False
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 52 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 192
On-line CPU(s) list: 0-191
Vendor ID: GenuineIntel
Model name: INTEL(R) XEON(R) PLATINUM 8558
CPU family: 6
Model: 207
Thread(s) per core: 2
Core(s) per socket: 48
Socket(s): 2
Stepping: 2
CPU max MHz: 4000.0000
CPU min MHz: 800.0000
BogoMIPS: 4200.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 4.5 MiB (96 instances)
L1i cache: 3 MiB (96 instances)
L2 cache: 192 MiB (96 instances)
L3 cache: 520 MiB (2 instances)
NUMA node(s): 4
NUMA node0 CPU(s): 0-23,96-119
NUMA node1 CPU(s): 24-47,120-143
NUMA node2 CPU(s): 48-71,144-167
NUMA node3 CPU(s): 72-95,168-191
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: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers
Vulnerability Spectre v2: Vulnerable; IBPB: disabled; STIBP: disabled; PBRSB-eIBRS: Vulnerable; BHI: Vulnerable
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Versions of relevant libraries:
[pip3] numpy==2.2.6
[pip3] nvidia-cublas-cu12==12.8.4.1
[pip3] nvidia-cuda-cupti-cu12==12.8.90
[pip3] nvidia-cuda-nvrtc-cu12==12.8.93
[pip3] nvidia-cuda-runtime-cu12==12.8.90
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cufft-cu12==11.3.3.83
[pip3] nvidia-curand-cu12==10.3.9.90
[pip3] nvidia-cusolver-cu12==11.7.3.90
[pip3] nvidia-cusparse-cu12==12.5.8.93
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.8.93
[pip3] nvidia-nvtx-cu12==12.8.90
[pip3] pytorch-triton==3.5.0+gitbbb06c03
[pip3] torch==2.10.0.dev20250926+cu128
[pip3] torchdata==0.11.0
[pip3] torchtitan==0.1.0.dev20250926+cu128
[conda] numpy 2.2.6 pypi_0 pypi
[conda] nvidia-cublas-cu12 12.8.4.1 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.8.90 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.8.93 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.8.90 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.10.2.21 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.3.3.83 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.9.90 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.7.3.90 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.5.8.93 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.7.1 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.27.5 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.8.93 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.8.90 pypi_0 pypi
[conda] pytorch-triton 3.5.0+gitbbb06c03 pypi_0 pypi
[conda] torch 2.10.0.dev20250926+cu128 pypi_0 pypi
[conda] torchdata 0.11.0 pypi_0 pypi
[conda] torchtitan 0.1.0.dev20250926+cu128 pypi_0 pypi
