{"id":23469118,"url":"https://github.com/hdelan/msc-hpc-final-project","last_synced_at":"2025-04-12T22:44:17.167Z","repository":{"id":170951419,"uuid":"377445628","full_name":"hdelan/MSc-HPC-Final-Project","owner":"hdelan","description":"In this project I implement a CUDA Lanczos method to approximate the matrix exponential. The matrix exponential is an important centrality measure for large, sparse graphs.","archived":false,"fork":false,"pushed_at":"2021-09-16T21:11:02.000Z","size":72006,"stargazers_count":1,"open_issues_count":1,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-12T22:44:14.333Z","etag":null,"topics":["cuda","graph-algorithms","krylov-methods"],"latest_commit_sha":null,"homepage":"","language":"Cuda","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/hdelan.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2021-06-16T09:42:37.000Z","updated_at":"2025-01-29T18:10:11.000Z","dependencies_parsed_at":null,"dependency_job_id":"c12027b7-3280-4ca3-98ab-f9e59e5b839c","html_url":"https://github.com/hdelan/MSc-HPC-Final-Project","commit_stats":null,"previous_names":["hdelan/msc-hpc-final-project"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hdelan%2FMSc-HPC-Final-Project","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hdelan%2FMSc-HPC-Final-Project/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hdelan%2FMSc-HPC-Final-Project/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hdelan%2FMSc-HPC-Final-Project/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hdelan","download_url":"https://codeload.github.com/hdelan/MSc-HPC-Final-Project/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248643048,"owners_count":21138353,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cuda","graph-algorithms","krylov-methods"],"created_at":"2024-12-24T14:59:35.994Z","updated_at":"2025-04-12T22:44:17.161Z","avatar_url":"https://github.com/hdelan.png","language":"Cuda","funding_links":[],"categories":[],"sub_categories":[],"readme":"# A CUDA Lanczos Method to Compute Graph Centrality\n\n\u003cfigure\u003e\n\u003cimg align=\"right\" src=\".images/Intro.png\" alt=\"First pic\"/\u003e  \n\u003c/figure\u003e\n\n\nIn the analysis of networks it is of great importance to identify central and non-central nodes. There are countless ways of measuring the centrality of nodes in a network, all with slightly different interpretations of what it means to be central. \n\nThe aim of this project is to compute the centrality of nodes in an undirected graph \u003cimg src=\"https://render.githubusercontent.com/render/math?math=A^T = A\"\u003e using the exponential matrix function: \n\n\u003cimg src=\"https://render.githubusercontent.com/render/math?math=f(A)x = e^Ax\"\u003e \n\n\u003cimg src=\"https://render.githubusercontent.com/render/math?math=e^Ax\"\u003e will be approximated without forming  \u003cimg src=\"https://render.githubusercontent.com/render/math?math=e^A\"\u003e explicitly, instead using Krylov subspace methods, most notably the Lanczos method. This makes computing \u003cimg src=\"https://render.githubusercontent.com/render/math?math=e^Ax\"\u003e scalable to the limits of the memory hardware, which are tested in this project. \n\nThis project computes \u003cimg src=\"https://render.githubusercontent.com/render/math?math=e^Ax\"\u003e both in serial and in parallel using CUDA.\n\n## Report\n\nThe accompanying report can be found at ```writeup.pdf```.\n\n## Data\nIn order to run code, data must be downloaded from:\n\nhttps://drive.google.com/drive/folders/1HdyMdnjphMtafk8TBbWc0L-2V4OeWlmj?usp=sharing\n\nand untarred. Please move the ```data/``` file (extracted from ```data.tar```) in this root directory.\n\n## Implementations\n\nThere are many implementations of the Lanczos method in this project directory:\n  - A serial method. See ```serial/README.md```\n  - A CUDA method using on-card multiplication. See ```parallel-mult-on-card/README.md```\n  - A CUDA method using two cards. See ```parallel-two-cards/README.md```\n  - A final CUDA implementation. Which uses a CUDA Lanczos method, and a multithreaded CBLAS multOut routine. See ```parallel-final/README.md```\n\nAll directories except ```final``` include tests that check the accuracy of the method and of kernels.\n\nEach implementation has its own ```lib/``` with slightly different code implementations. The final version of code can be found in ```parallel-final/lib```.\n\nAll code found in ```parallel-*``` directories must be run on a CUDA-enabled device.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhdelan%2Fmsc-hpc-final-project","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhdelan%2Fmsc-hpc-final-project","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhdelan%2Fmsc-hpc-final-project/lists"}