About the inference speed.

[sanfutian]

Thank you for your great work. I encountered an issue while trying to reproduce this project. In the paper, the inference speed is stated as 200 fps, but during my testing, it's only around 10+ fps, with VoxelNet consuming a significant amount of time. I have installed all the required dependencies as instructed. Below is my system environment:

---

System environment: sys.platform: linux Python: 3.9.19 (main, May 6 2024, 19:43:03) [GCC 11.2.0] CUDA available: True numpy_random_seed: 482379611 GPU 0,1,2,3,4,5,6,7: NVIDIA GeForce RTX 3090 CUDA_HOME: /usr/local/cuda-11.8 NVCC: Cuda compilation tools, release 11.8, V11.8.89 GCC: gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 PyTorch: 2.0.1 PyTorch compiling details: PyTorch built with:

• GCC 9.3
• C++ Version: 201703
• Intel(R) oneAPI Math Kernel Library Version 2023.1-Product Build 20230303 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.8
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_37,code=compute_37
• CuDNN 8.7
• Magma 2.6.1

• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.8, CUDNN_VERSION=8.7.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.0.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF,

  TorchVision: 0.15.2 OpenCV: 4.9.0 MMEngine: 0.7.4

Runtime environment: cudnn_benchmark: False mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0} dist_cfg: {'backend': 'nccl'} seed: 482379611 Distributed launcher: none Distributed training: False GPU number: 1

[xmm-prio]

Do you install the dependency of spconv?

[sanfutian]

yes, I install "spconv-cu118 2.3.6"

[xmm-prio]

What's your speed test setup like? The network and data on the gpu?

[sanfutian]

Thank you for responding so quickly! It seems like the previous inference speed was indeed affected by other programs running on the server. In my recent test on an NVIDIA GTX 3090 GPU, the latest inference speeds are as follows: bevtrack(VoxelNet+BMM+head): 14.3ms/frame; pre/postprocess: 7.8ms/frame. It's great to hear that the performance of VoxelNet has significantly improved. Please note that the server is not idle due to the activities of my colleagues' programs. I'll update the latest inference speed under idle conditions.

[xmm-prio]

Wishing you all the best and hoping that BEVTrack could help your further research.