[Bug]: TTFT latency issue with Qwen3.5-35B-A3B model using vllm #35625
Description
Your current environment
The output of python collect_env.py
==============================
System Info
==============================
OS : Ubuntu 24.04.3 LTS (x86_64)
GCC version : (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0
Clang version : Could not collect
CMake version : version 4.2.1
Libc version : glibc-2.39
==============================
PyTorch Info
==============================
PyTorch version : 2.11.0+cu130
Is debug build : False
CUDA used to build PyTorch : 13.0
ROCM used to build PyTorch : N/A
==============================
Python Environment
==============================
Python version : 3.12.3 (main, Jan 22 2026, 20:57:42) [GCC 13.3.0] (64-bit runtime)
Python platform : Linux-6.17.0-14-generic-x86_64-with-glibc2.39
==============================
CUDA / GPU Info
==============================
Is CUDA available : True
CUDA runtime version : Could not collect
CUDA_MODULE_LOADING set to :
GPU models and configuration :
GPU 0: NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
GPU 1: NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
Nvidia driver version : 580.126.09
cuDNN version : Could not collect
HIP runtime version : N/A
MIOpen runtime version : N/A
Is XNNPACK available : True
==============================
CPU Info
==============================
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): 48
On-line CPU(s) list: 0-47
Vendor ID: AuthenticAMD
Model name: AMD Ryzen Threadripper 9960X 24-Cores
CPU family: 26
Model: 8
Thread(s) per core: 2
Core(s) per socket: 24
Socket(s): 1
Stepping: 1
Frequency boost: enabled
CPU(s) scaling MHz: 30%
CPU max MHz: 5489.7642
CPU min MHz: 1223.6230
BogoMIPS: 8387.92
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 amd_lbr_v2 nopl xtopology nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic 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 cpuid_fault cpb cat_l3 cdp_l3 hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local user_shstk avx_vnni avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl vnmi avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid bus_lock_detect movdiri movdir64b overflow_recov succor smca fsrm avx512_vp2intersect flush_l1d debug_swap amd_lbr_pmc_freeze
Virtualization: AMD-V
L1d cache: 1.1 MiB (24 instances)
L1i cache: 768 KiB (24 instances)
L2 cache: 24 MiB (24 instances)
L3 cache: 128 MiB (4 instances)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Gather data sampling: Not affected
Vulnerability Ghostwrite: 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 Old microcode: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Mitigation; Reduced Speculation
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; Enhanced / Automatic IBRS; IBPB conditional; STIBP always-on; 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] flashinfer-python==0.6.4
[pip3] numpy==2.4.2
[pip3] nvidia-cublas==13.2.1.1
[pip3] nvidia-cuda-cupti==13.0.85
[pip3] nvidia-cuda-nvrtc==13.0.88
[pip3] nvidia-cuda-runtime==13.0.96
[pip3] nvidia-cudnn-cu13==9.19.1.2
[pip3] nvidia-cudnn-frontend==1.18.0
[pip3] nvidia-cufft==12.0.0.61
[pip3] nvidia-cufile==1.15.1.6
[pip3] nvidia-curand==10.4.0.35
[pip3] nvidia-cusolver==12.0.4.66
[pip3] nvidia-cusparse==12.6.3.3
[pip3] nvidia-cusparselt-cu13==0.8.0
[pip3] nvidia-cutlass-dsl==4.4.0
[pip3] nvidia-cutlass-dsl-libs-base==4.4.0
[pip3] nvidia-ml-py==13.590.48
[pip3] nvidia-nccl-cu13==2.28.9
[pip3] nvidia-nvjitlink==13.0.88
[pip3] nvidia-nvshmem-cu13==3.4.5
[pip3] nvidia-nvtx==13.0.85
[pip3] pyzmq==27.1.0
[pip3] torch==2.11.0+cu130
[pip3] torchvision==0.26.0+cu130
[pip3] transformers==4.57.6
[pip3] triton==3.6.0
[conda] Could not collect
==============================
vLLM Info
==============================
ROCM Version : Could not collect
vLLM Version : 0.16.1rc1.dev12+g111d86906.d20260226 (git sha: 111d86906, date: 20260226)
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled
GPU Topology:
GPU0 GPU1 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NODE 0-47 0 N/A
GPU1 NODE X 0-47 0 N/A
Legend:
X = Self
SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI)
NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node
PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU)
PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge)
PIX = Connection traversing at most a single PCIe bridge
NV# = Connection traversing a bonded set of # NVLinks
==============================
Environment Variables
==============================
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_admin2
🐛 Describe the bug
Hello, I am experiencing severe tail latency for TTFT when running a concurrency benchmark on a Blackwell 6000 Pro and DGX Spark
Reproduce command:
vllm serve Qwen/Qwen3.5-35B-A3B \
--port 8000 \
--tensor-parallel-size 1 \
--max-model-len 262144 \
--reasoning-parser qwen3 \
--speculative-config '{"method":"qwen3_next_mtp","num_speculative_tokens":2}' \
--gpu-memory-utilization 0.8 \
--language-model-only
Output:
============ Serving Benchmark Result ============
Successful requests: 1000
Failed requests: 0
Maximum request concurrency: 100
Benchmark duration (s): 38.30
Total input tokens: 128000
Total generated tokens: 128000
Request throughput (req/s): 26.11
Output token throughput (tok/s): 3341.89
Peak output token throughput (tok/s): 2100.00
Peak concurrent requests: 161.00
Total token throughput (tok/s): 6683.78
---------------Time to First Token----------------
Mean TTFT (ms): 611.98
Median TTFT (ms): 191.57
P99 TTFT (ms): 4756.51
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms): 24.27
Median TPOT (ms): 22.93
P99 TPOT (ms): 49.27
---------------Inter-token Latency----------------
Mean ITL (ms): 58.46
Median ITL (ms): 57.57
P99 ITL (ms): 78.27
---------------Speculative Decoding---------------
Acceptance rate (%): 71.28
Acceptance length: 2.43
Drafts: 52726
Draft tokens: 105452
Accepted tokens: 75168
Per-position acceptance (%):
Position 0: 76.73
Position 1: 65.84
==================================================
DGX Spark:
============ Serving Benchmark Result ============
Successful requests: 1000
Failed requests: 0
Maximum request concurrency: 100
Benchmark duration (s): 296.70
Total input tokens: 128000
Total generated tokens: 128000
Request throughput (req/s): 3.37
Output token throughput (tok/s): 431.42
Peak output token throughput (tok/s): 300.00
Peak concurrent requests: 114.00
Total token throughput (tok/s): 862.84
---------------Time to First Token----------------
Mean TTFT (ms): 2242.31
Median TTFT (ms): 1550.75
P99 TTFT (ms): 10033.63
-----Time per Output Token (excl. 1st token)------
Mean TPOT (ms): 206.55
Median TPOT (ms): 194.25
P99 TPOT (ms): 423.11
---------------Inter-token Latency----------------
Mean ITL (ms): 493.15
Median ITL (ms): 496.39
P99 ITL (ms): 640.49
---------------Speculative Decoding---------------
Acceptance rate (%): 70.24
Acceptance length: 2.40
Drafts: 53194
Draft tokens: 106388
Accepted tokens: 74729
Per-position acceptance (%):
Position 0: 75.77
Position 1: 64.71
==================================================
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.