search

rapidsai / cugraph

cuGraph - RAPIDS Graph Analytics Library
https://docs.rapids.ai/api/cugraph/stable/
Apache License 2.0
1.69k stars 300 forks source link

Algorithm - Closeness Centrality #4302

Open ChuckHastings opened 6 months ago

ChuckHastings commented 6 months ago

Implement Closeness Centrality

Fre0Grella commented 5 months ago

Hi, I'd like to fix to this issue. I'm a student from the univesity of bologna and for my thesis project I want to implement the GPU-acellerated version of this algorithm.

ChuckHastings commented 5 months ago

We would welcome an implementation of Closeness Centrality!

Happy to collaborate and help you understand anything in the cugraph framework. You might start with looking at our implementation of Betweenness Centrality, which will probably have some similarities with an implementation of Closeness Centrality.

Fre0Grella commented 4 months ago

Hi, i'm trying to work with the university cluster, and i'm having problem to build the project from the script ./build.sh This is the error:

Building for the architecture of the GPU in the system... CMake Error at /public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:814 (message): Compiling the CUDA compiler identification source file "CMakeCUDACompilerId.cu" failed.

Compiler: /public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/bin/nvcc

Build flags:

Id flags: --keep;--keep-dir;tmp -v

The output was:

1

$ _NVVMBRANCH=nvvm

$ SPACE=

$ CUDART=cudart

$ HERE=/usr/lib/nvidia-cuda-toolkit/bin

$ THERE=/usr/lib/nvidia-cuda-toolkit/bin

$ _TARGETSIZE=

$ _TARGETDIR=

$ _TARGETSIZE=64

$ NVVMIR_LIBRARY_DIR=/usr/lib/nvidia-cuda-toolkit/libdevice

$

PATH=/usr/lib/nvidia-cuda-toolkit/bin:/public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/bin:/public.hpc/marco.galeri/miniforge3/condabin:/home/students/marco.galeri/.vscode-server/cli/servers/Stable-dc96b837cf6bb4af9cd736aa3af08cf8279f7685/server/bin/remote-cli:/usr/local/bin:/usr/bin:/bin:/usr/games

$ LIBRARIES= -L/usr/lib/x86_64-linux-gnu/stubs -L/usr/lib/x86_64-linux-gnu

$ rm tmp/a_dlink.reg.c

$

"/public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/bin"/x86_64-conda-linux-gnu-c++ -DCUDA_ARCH__=520 -E -x c++ -DCUDA_DOUBLE_MATH_FUNCTIONS -DCUDACC -DNVCC -DCUDACC_VER_MAJOR=11 -DCUDACC_VER_MINOR=2 -DCUDACC_VER_BUILD=152 -DCUDA_API_VER_MAJOR=11 -DCUDA_API_VER_MINOR__=2 -include "cuda_runtime.h" -m64 "CMakeCUDACompilerId.cu" -o "tmp/CMakeCUDACompilerId.cpp1.ii"

: fatal error: cuda_runtime.h: No such file or directory compilation terminated. # --error 0x1 -- Call Stack (most recent call first): /public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:8 (CMAKE_DETERMINE_COMPILER_ID_BUILD) /public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/share/cmake-3.29/Modules/CMakeDetermineCompilerId.cmake:53 (__determine_compiler_id_test) /public.hpc/marco.galeri/miniforge3/envs/cugraph_dev/share/cmake-3.29/Modules/CMakeDetermineCUDACompiler.cmake:131 (CMAKE_DETERMINE_COMPILER_ID) CMakeLists.txt:28 (project) -- Configuring incomplete, errors occurred!
ChuckHastings commented 4 months ago

Can you provide output from nvidia-smi and nvcc--version?

Fre0Grella commented 4 months ago

nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2021 NVIDIA Corporation Built on Sun_Feb_14_21:12:58_PST_2021 Cuda compilation tools, release 11.2, V11.2.152 Build cuda_11.2.r11.2/compiler.29618528_0 Tue May 21 08:39:13 2024
+-----------------------------------------------------------------------------+ | NVIDIA-SMI 470.182.03 Driver Version: 470.182.03 CUDA Version: 11.4 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA GeForce ... On | 00000000:00:10.0 Off | N/A | | 86% 81C P0 122W / 250W | 1MiB / 11019MiB | 99% Default | | | | N/A | +-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | No running processes found | +-----------------------------------------------------------------------------+

Fre0Grella commented 2 months ago

I'm not prepared enough to fix this issue yet, i drop this project

ChuckHastings commented 2 months ago

Thanks for your interest. If you have further interest in the future please let us know.