La fonctionnalité que nous sommes en train d'ajouter à cuGraph, c'est le support du paramètre prevent_overlapping, pour produire de belles visualisations de graphes où les noeuds sont correctement espacés.
Ça existe déjà sur Gephi, mais cuGraph permettra de traiter des données bien plus massives !
Ça existe déjà sur Gephi, mais cuGraph permettra de traiter des données bien plus massives !
May 12, 2025 at 11:51 AM
La fonctionnalité que nous sommes en train d'ajouter à cuGraph, c'est le support du paramètre prevent_overlapping, pour produire de belles visualisations de graphes où les noeuds sont correctement espacés.
Ça existe déjà sur Gephi, mais cuGraph permettra de traiter des données bien plus massives !
Ça existe déjà sur Gephi, mais cuGraph permettra de traiter des données bien plus massives !
Le truc de bosser avec quelqu'un qui est passionnée de mathématique compliqué :
Il est en stream et parle d'un problème qu'on a avec Agoratlas...
Et il propose de faire une contrib publique à une énorme bibliothèque opensource pour y répondre 😆
(Ajout de fonctionnalité sur CuGraph, si vous voyez)
Il est en stream et parle d'un problème qu'on a avec Agoratlas...
Et il propose de faire une contrib publique à une énorme bibliothèque opensource pour y répondre 😆
(Ajout de fonctionnalité sur CuGraph, si vous voyez)
Est-ce que ça vous intéresse un petit stream cet après-midi pour vous montrer mon projet du moment avec @agoratlas.com ?
C'est un moteur de visualisation de graphes en Python pour @flefgraph.bsky.social 😊
C'est un moteur de visualisation de graphes en Python pour @flefgraph.bsky.social 😊
April 8, 2025 at 12:28 PM
Le truc de bosser avec quelqu'un qui est passionnée de mathématique compliqué :
Il est en stream et parle d'un problème qu'on a avec Agoratlas...
Et il propose de faire une contrib publique à une énorme bibliothèque opensource pour y répondre 😆
(Ajout de fonctionnalité sur CuGraph, si vous voyez)
Il est en stream et parle d'un problème qu'on a avec Agoratlas...
Et il propose de faire une contrib publique à une énorme bibliothèque opensource pour y répondre 😆
(Ajout de fonctionnalité sur CuGraph, si vous voyez)
Bit of good news. I picked up two sponsors for my Python YouTube channel: Dell and NVIDIA. I'll be making a few series showcasing NVIDIA's libraries: cuDF, cuGraph, and cuML. Anyone have suggestions for a large humanities dataset for the cuDF or cuGraph videos/notebooks? I will have 32 GB of VRAM.
December 29, 2024 at 4:13 PM
Bit of good news. I picked up two sponsors for my Python YouTube channel: Dell and NVIDIA. I'll be making a few series showcasing NVIDIA's libraries: cuDF, cuGraph, and cuML. Anyone have suggestions for a large humanities dataset for the cuDF or cuGraph videos/notebooks? I will have 32 GB of VRAM.
🌐 Tutorial Spotlight: Ever wondered how to find hidden patterns in messy, connected data? @ericmjl.bsky.social teaches NetworkX fundamentals, then dives into LLMs, cuGraph, and more. Perfect for beginners and curious pros alike 🧠 #SciPy2026 🔗 scipy2026.scipy.org
June 5, 2026 at 1:00 AM
🌐 Tutorial Spotlight: Ever wondered how to find hidden patterns in messy, connected data? @ericmjl.bsky.social teaches NetworkX fundamentals, then dives into LLMs, cuGraph, and more. Perfect for beginners and curious pros alike 🧠 #SciPy2026 🔗 scipy2026.scipy.org
Thanks, NVIDIA and Dell for sponsoring a series of videos on my channel. This will the first in many videos that cover working with big data vai CUDA-Accelerated Python packages, like cuDF, cuGraph, and cuML.
Video: youtu.be/7zxUnXHhdfc
#MachineLearning
Video: youtu.be/7zxUnXHhdfc
#MachineLearning
What is Big Data and Scalability
YouTube video by Python Tutorials for Digital Humanities
youtu.be
February 3, 2025 at 1:11 PM
Thanks, NVIDIA and Dell for sponsoring a series of videos on my channel. This will the first in many videos that cover working with big data vai CUDA-Accelerated Python packages, like cuDF, cuGraph, and cuML.
Video: youtu.be/7zxUnXHhdfc
#MachineLearning
Video: youtu.be/7zxUnXHhdfc
#MachineLearning
Notre deuxième projet R&D du moment est cuGraph, une bibliothèque d'algorithmes de graphes développée par NVIDIA pour du calcul GPU haute performance.
À la base, c'était simplement pour ajouter une nouvelle fonctionnalité, mais au passage on a trouvé (et réparé) des bugs qui avaient plus de 6 ans !
À la base, c'était simplement pour ajouter une nouvelle fonctionnalité, mais au passage on a trouvé (et réparé) des bugs qui avaient plus de 6 ans !
May 12, 2025 at 11:51 AM
Notre deuxième projet R&D du moment est cuGraph, une bibliothèque d'algorithmes de graphes développée par NVIDIA pour du calcul GPU haute performance.
À la base, c'était simplement pour ajouter une nouvelle fonctionnalité, mais au passage on a trouvé (et réparé) des bugs qui avaient plus de 6 ans !
À la base, c'était simplement pour ajouter une nouvelle fonctionnalité, mais au passage on a trouvé (et réparé) des bugs qui avaient plus de 6 ans !
developer.nvidia.com/blog/network... HUGE news!
October 29, 2024 at 5:03 AM
developer.nvidia.com/blog/network... HUGE news!
Update : le boulot est moins simple que prévu (on doit d'abord rattraper quelques étourderies des ingénieurs Nvidia -- github.com/rapidsai/cug... 😇), mais ça avance !
La fonctionnalité arrive dans les prochains mois, à disposition pour échanger si vous voulez l'intégrer avant que ce soit public :)
La fonctionnalité arrive dans les prochains mois, à disposition pour échanger si vous voulez l'intégrer avant que ce soit public :)
ForceAtlas2 threads returning prematurely? · Issue #5029 · rapidsai/cugraph
I'm currently working on cuGraph's FA2 implementation to add support for the prevent_overlapping parameter (#5022 ), which is documented but not supported. I found something confusing in the implem...
github.com
April 14, 2025 at 8:44 PM
Update : le boulot est moins simple que prévu (on doit d'abord rattraper quelques étourderies des ingénieurs Nvidia -- github.com/rapidsai/cug... 😇), mais ça avance !
La fonctionnalité arrive dans les prochains mois, à disposition pour échanger si vous voulez l'intégrer avant que ce soit public :)
La fonctionnalité arrive dans les prochains mois, à disposition pour échanger si vous voulez l'intégrer avant que ce soit public :)
Jarrod Millman is a core contributor to the NetworkX that is used to understand human cells, the dynamics of beliefs in social networks, and power grid failures (among many other applications!)
The team is working with NVIDIA to give users faster calculations with “no code” changes required.
The team is working with NVIDIA to give users faster calculations with “no code” changes required.
March 13, 2025 at 5:01 PM
Jarrod Millman is a core contributor to the NetworkX that is used to understand human cells, the dynamics of beliefs in social networks, and power grid failures (among many other applications!)
The team is working with NVIDIA to give users faster calculations with “no code” changes required.
The team is working with NVIDIA to give users faster calculations with “no code” changes required.
adbcug-adapter 2.0.2
Convert ArangoDB graphs to cuGraph & vice-versa.
Author: Anthony Mahanna
🏠Homepage
Convert ArangoDB graphs to cuGraph & vice-versa.
Author: Anthony Mahanna
🏠Homepage
January 24, 2025 at 7:01 AM
adbcug-adapter 2.0.2
Convert ArangoDB graphs to cuGraph & vice-versa.
Author: Anthony Mahanna
🏠Homepage
Convert ArangoDB graphs to cuGraph & vice-versa.
Author: Anthony Mahanna
🏠Homepage
Oh nice! Would you mind letting me know when you use cuGraph how well it works from you and some things you noticed?
May 2, 2025 at 3:01 PM
Oh nice! Would you mind letting me know when you use cuGraph how well it works from you and some things you noticed?
Anyone working in #graph analytics knows that NetworkX is super popular for prototyping, but it can get quite slow. This blog introduces zero-code-change GPU acceleration for NetworkX code with massive speedups. (Think 100x or more for large graphs.): developer.nvidia.com/blog/network...
NetworkX Introduces Zero Code Change Acceleration Using NVIDIA cuGraph | NVIDIA Technical Blog
NetworkX accelerated by NVIDIA cuGraph is a newly released backend co-developed with the NetworkX team. NVIDIA cuGraph provides GPU acceleration for popular graph algorithms such as PageRank, Louvain…
developer.nvidia.com
November 14, 2024 at 6:59 PM
Anyone working in #graph analytics knows that NetworkX is super popular for prototyping, but it can get quite slow. This blog introduces zero-code-change GPU acceleration for NetworkX code with massive speedups. (Think 100x or more for large graphs.): developer.nvidia.com/blog/network...
OK, will do. Exciting to hear about cuGraph will check it out.
Until now I have been using igraph.org - but the igraph indices need to be 32bit ints, whereas my graph has 64bit IDs making conversions a pain in the proverbial.
Until now I have been using igraph.org - but the igraph indices need to be 32bit ints, whereas my graph has 64bit IDs making conversions a pain in the proverbial.
May 2, 2025 at 11:23 AM
OK, will do. Exciting to hear about cuGraph will check it out.
Until now I have been using igraph.org - but the igraph indices need to be 32bit ints, whereas my graph has 64bit IDs making conversions a pain in the proverbial.
Until now I have been using igraph.org - but the igraph indices need to be 32bit ints, whereas my graph has 64bit IDs making conversions a pain in the proverbial.
Two GPU tools for data too big to look at
cuGraph does graph analytics on the GPU; Omniverse does real-time 3D simulation.
#NCAAIIO #AI #DataScience #GPUComputing
Full NCA-AIIO explanation, free: https://navyduck.com/nvidia/ai-infrastructure/nca-aiio/q117-a-research-team-needs-to
cuGraph does graph analytics on the GPU; Omniverse does real-time 3D simulation.
#NCAAIIO #AI #DataScience #GPUComputing
Full NCA-AIIO explanation, free: https://navyduck.com/nvidia/ai-infrastructure/nca-aiio/q117-a-research-team-needs-to
September 11, 2026 at 5:00 PM
Two GPU tools for data too big to look at
cuGraph does graph analytics on the GPU; Omniverse does real-time 3D simulation.
#NCAAIIO #AI #DataScience #GPUComputing
Full NCA-AIIO explanation, free: https://navyduck.com/nvidia/ai-infrastructure/nca-aiio/q117-a-research-team-needs-to
cuGraph does graph analytics on the GPU; Omniverse does real-time 3D simulation.
#NCAAIIO #AI #DataScience #GPUComputing
Full NCA-AIIO explanation, free: https://navyduck.com/nvidia/ai-infrastructure/nca-aiio/q117-a-research-team-needs-to
Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph
Nelson Aloysio Reis de Almeida Passos, Emanuele Carlini, Salvatore Trani
#arXiv #cs.DC #cs.LG
Nelson Aloysio Reis de Almeida Passos, Emanuele Carlini, Salvatore Trani
#arXiv #cs.DC #cs.LG
Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph
This work addresses community detection in temporal networks through GPU-accelerated extensions of spectral clustering and modularity-based algorithms originally designed for static graphs. Built on the NVIDIA RAPIDS ecosystem, the framework enables the characterization and tracking of communities …
arxiv.org
August 5, 2026 at 6:48 PM
🤖: Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph - 2608.03695v1 Die 3 wichtigsten Erkenntnisse -Geometrische Grenze & Semantische Interferenz:*Eine verlässliche Speicherung von Fakten erfordert im Vektorraum zueinander orthogonal stehende Schlüssel. Da neuronale Netze
August 5, 2026 at 5:00 PM
🤖: Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph - 2608.03695v1 Die 3 wichtigsten Erkenntnisse -Geometrische Grenze & Semantische Interferenz:*Eine verlässliche Speicherung von Fakten erfordert im Vektorraum zueinander orthogonal stehende Schlüssel. Da neuronale Netze
Nelson Aloysio Reis de Almeida Passos, Emanuele Carlini, Salvatore Trani: Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph https://arxiv.org/abs/2608.03695 https://arxiv.org/pdf/2608.03695 https://arxiv.org/html/2608.03695
August 5, 2026 at 6:40 AM
Nelson Aloysio Reis de Almeida Passos, Emanuele Carlini, Salvatore Trani: Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph https://arxiv.org/abs/2608.03695 https://arxiv.org/pdf/2608.03695 https://arxiv.org/html/2608.03695
Breaking News! @NVIDIA selects @Prof_DavidBader @GeorgiaTech to join #NVAIL Program with Focus on #Graph #Analytics
@RAPIDSai #cuGraph @NvidiaAI #GPU #DataScience http://bit.ly/2VRcdsC
@RAPIDSai #cuGraph @NvidiaAI #GPU #DataScience http://bit.ly/2VRcdsC
November 23, 2024 at 3:23 PM
Breaking News! @NVIDIA selects @Prof_DavidBader @GeorgiaTech to join #NVAIL Program with Focus on #Graph #Analytics
@RAPIDSai #cuGraph @NvidiaAI #GPU #DataScience http://bit.ly/2VRcdsC
@RAPIDSai #cuGraph @NvidiaAI #GPU #DataScience http://bit.ly/2VRcdsC
Achieve 100x Speedups in Graph Analytics Using Nx-cugraph
NetworkX is a powerhouse for graph analytics in Python, beloved for its ease of use and vast community. As graphs grow, its pure-Python nature can lead to performance bottlenecks. Enter `nx-cugraph`, a RAPIDS back…
#hackernews #news #nvidia
NetworkX is a powerhouse for graph analytics in Python, beloved for its ease of use and vast community. As graphs grow, its pure-Python nature can lead to performance bottlenecks. Enter `nx-cugraph`, a RAPIDS back…
#hackernews #news #nvidia
Achieve 100x Speedups in Graph Analytics Using Nx-cugraph
NetworkX is a powerhouse for graph analytics in Python, beloved for its ease of use and vast community. As graphs grow, its pure-Python nature can lead to performance bottlenecks. Enter `nx-cugraph`, a RAPIDS backend that lets NetworkX leverage the power of NVIDIA GPUs.
hackernoon.com
May 29, 2025 at 4:16 AM
Achieve 100x Speedups in Graph Analytics Using Nx-cugraph
NetworkX is a powerhouse for graph analytics in Python, beloved for its ease of use and vast community. As graphs grow, its pure-Python nature can lead to performance bottlenecks. Enter `nx-cugraph`, a RAPIDS back…
#hackernews #news #nvidia
NetworkX is a powerhouse for graph analytics in Python, beloved for its ease of use and vast community. As graphs grow, its pure-Python nature can lead to performance bottlenecks. Enter `nx-cugraph`, a RAPIDS back…
#hackernews #news #nvidia
We have integrated the ForceAtlas2 layout algorithm from cugraph. This GPU-based implementation allows Communalytic to render and visualize significantly larger graphs in a fraction of the time.
June 25, 2026 at 11:01 AM
We have integrated the ForceAtlas2 layout algorithm from cugraph. This GPU-based implementation allows Communalytic to render and visualize significantly larger graphs in a fraction of the time.
That is a really cool news! Nvidia presented an accelerator for NetworkX, the #Python library for working with graph data. Without changing any part of the code, one can achieve 50-500x performance boost with GPU over CPU.
nvda.ws/3YyDO4A
nvda.ws/3YyDO4A
NetworkX Introduces Zero Code Change Acceleration Using NVIDIA cuGraph | NVIDIA Technical Blog
NetworkX accelerated by NVIDIA cuGraph is a newly released backend co-developed with the NetworkX team. NVIDIA cuGraph provides GPU acceleration for popular graph algorithms such as PageRank, Louvain…
nvda.ws
October 23, 2024 at 7:25 AM
That is a really cool news! Nvidia presented an accelerator for NetworkX, the #Python library for working with graph data. Without changing any part of the code, one can achieve 50-500x performance boost with GPU over CPU.
nvda.ws/3YyDO4A
nvda.ws/3YyDO4A
Ahh hmm yeah that would be problematic. If you already have a Python backend and a GPU then cuGraph may work really well for you. Is this for Iconclass?
May 2, 2025 at 11:29 AM
Ahh hmm yeah that would be problematic. If you already have a Python backend and a GPU then cuGraph may work really well for you. Is this for Iconclass?
Actually, for graph traversal you can either wait for my upcoming video or just use it now: cuGraph. NVIDIA is sponsoring three series on my channel: cuDF, cuGraph, and cuML. cuGraph has most of the algos (and adding more) from networkx.
May 2, 2025 at 11:15 AM
Actually, for graph traversal you can either wait for my upcoming video or just use it now: cuGraph. NVIDIA is sponsoring three series on my channel: cuDF, cuGraph, and cuML. cuGraph has most of the algos (and adding more) from networkx.