{"id":13439795,"url":"https://github.com/RRZE-HPC/likwid","last_synced_at":"2025-03-20T09:30:48.353Z","repository":{"id":31481121,"uuid":"35045238","full_name":"RRZE-HPC/likwid","owner":"RRZE-HPC","description":"Performance monitoring and benchmarking suite","archived":false,"fork":false,"pushed_at":"2024-10-17T08:49:11.000Z","size":32034,"stargazers_count":1663,"open_issues_count":68,"forks_count":227,"subscribers_count":66,"default_branch":"master","last_synced_at":"2024-10-19T17:20:06.482Z","etag":null,"topics":["amd-gpu","armv8","assembly","benchmarking","c","hardware-performance-counters","hwloc","instrumentation","likwid","linux","lua","nvidia-gpu","performance-analysis","performance-engineering","pin","power9","profiling","threading","x86"],"latest_commit_sha":null,"homepage":"https://hpc.fau.de/research/tools/likwid/","language":"C","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/RRZE-HPC.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG","contributing":null,"funding":null,"license":"COPYING","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":"2015-05-04T16:13:45.000Z","updated_at":"2024-10-18T00:14:38.000Z","dependencies_parsed_at":"2023-12-02T02:15:01.505Z","dependency_job_id":"5cc1c3c9-3c16-4292-b7f5-ee841c76292d","html_url":"https://github.com/RRZE-HPC/likwid","commit_stats":{"total_commits":2653,"total_committers":71,"mean_commits":37.36619718309859,"dds":0.5974368639276291,"last_synced_commit":"d686eabcde3bb046b9061aac5325dd0ded009e8e"},"previous_names":["rrze-likwid/likwid"],"tags_count":32,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RRZE-HPC%2Flikwid","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RRZE-HPC%2Flikwid/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RRZE-HPC%2Flikwid/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RRZE-HPC%2Flikwid/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/RRZE-HPC","download_url":"https://codeload.github.com/RRZE-HPC/likwid/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":221745184,"owners_count":16873733,"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":["amd-gpu","armv8","assembly","benchmarking","c","hardware-performance-counters","hwloc","instrumentation","likwid","linux","lua","nvidia-gpu","performance-analysis","performance-engineering","pin","power9","profiling","threading","x86"],"created_at":"2024-07-31T03:01:17.154Z","updated_at":"2025-03-20T09:30:48.347Z","avatar_url":"https://github.com/RRZE-HPC.png","language":"C","readme":"--------------------------------------------------------------------------------\nIntroduction\n--------------------------------------------------------------------------------\n\nLikwid is a simple to install and use toolsuite of command line applications and a library\nfor performance oriented programmers. It works for Intel, AMD, ARMv8 and POWER9\nprocessors on the Linux operating system. There is additional support for Nvidia and AMD GPUs.\nThere is support for ARMv7 and POWER8/9 but there is currently no test machine in\nour hands to test them properly.\n\n[LIKWID Playlist (YouTube)](https://www.youtube.com/playlist?list=PLxVedhmuwLq2CqJpAABDMbZG8Whi7pKsk)\n\n[![Build Status](https://gitos.rrze.fau.de/ub55yzis/likwid/badges/master/pipeline.svg)](https://gitos.rrze.fau.de/ub55yzis/likwid/-/commits/master) [![General LIKWID DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.4275676.svg)](https://doi.org/10.5281/zenodo.4275676)\n\nIt consists of:\n\n- likwid-topology: print thread, cache and NUMA topology\n- likwid-perfctr: configure and read out hardware performance counters on Intel, AMD, ARM and POWER processors and Nvidia GPUs\n- likwid-powermeter: read out RAPL Energy information and get info about Turbo mode steps\n- likwid-pin: pin your threaded application (pthread, Intel and gcc OpenMP to dedicated processors)\n- likwid-bench: Micro benchmarking platform for CPU architectures\n- likwid-features: Print and manipulate cpu features like hardware prefetchers (x86 only)\n- likwid-genTopoCfg: Dumps topology information to a file\n- likwid-mpirun: Wrapper to start MPI and Hybrid MPI/OpenMP applications (Supports Intel MPI, OpenMPI, MPICH and SLURM)\n- likwid-perfscope: Frontend to the timeline mode of likwid-perfctr, plots live graphs of performance metrics using gnuplot\n- likwid-memsweeper: Sweep memory of NUMA domains and evict cachelines from the last level cache\n- likwid-setFrequencies: Tool to control the CPU and Uncore frequencies (x86 only)\n- likwid-sysFeatures: Tool to system settings like frequencies, powercaps and prefetchers (experimental)\n\nFor further information please take a look at the [Wiki](https://github.com/RRZE-HPC/likwid/wiki) or contact us via Matrix chat [LIKWID General](https://matrix.to/#/#likwid:matrix.org?via=matrix.org).\n\n\n--------------------------------------------------------------------------------\nSupported architectures\n--------------------------------------------------------------------------------\nIntel\n- Intel Atom\n- Intel Pentium M\n- Intel Core2\n- Intel Nehalem\n- Intel NehalemEX\n- Intel Westmere\n- Intel WestmereEX\n- Intel Xeon Phi (KNC)\n- Intel Silvermont \u0026 Airmont\n- Intel Goldmont\n- Intel SandyBridge\n- Intel SandyBridge EP/EN\n- Intel IvyBridge\n- Intel IvyBridge EP/EN/EX\n- Intel Xeon Phi (KNL, KNM)\n- Intel Haswell\n- Intel Haswell EP/EN/EX\n- Intel Broadwell\n- Intel Broadwell D\n- Intel Broadwell EP\n- Intel Skylake\n- Intel Kabylake\n- Intel Coffeelake\n- Intel Skylake SP\n- Intel Cascadelake SP\n- Intel Icelake\n- Intel Icelake SP\n- Intel Tigerlake (experimental)\n- Intel SapphireRapids\n- Intel EmeraldRapids\n\nAMD\n- AMD K8\n- AMD K10\n- AMD Interlagos\n- AMD Kabini\n- AMD Zen\n- AMD Zen2\n- AMD Zen3\n- AMD Zen4\n\nARM\n- ARMv7\n- ARMv8\n- Special support for Marvell Thunder X2\n- Fujitsu A64FX\n- ARM Neoverse N1 (AWS Graviton 2)\n- ARM Neoverse V1\n- HiSilicon TSV110\n- Apple M1 (only with Linux)\n\nPOWER (experimental)\n- IBM POWER8\n- IBM POWER9\n\nNvidia GPUs\n\nAMD GPUs\n\n--------------------------------------------------------------------------------\nDownload, Build and Install\n--------------------------------------------------------------------------------\nYou can get the releases of LIKWID at:\nhttp://ftp.fau.de/pub/likwid/\n\nFor build and installation hints see INSTALL file or check the build instructions\npage in the wiki https://github.com/RRZE-HPC/likwid/wiki/Build\n\nFor quick install:\n```bash\nVERSION=stable\nwget http://ftp.fau.de/pub/likwid/likwid-$VERSION.tar.gz\ntar -xaf likwid-$VERSION.tar.gz\ncd likwid-*\nvi config.mk # configure build, e.g. change installation prefix and architecture flags\nmake\nsudo make install # sudo required to install the access daemon with proper permissions\n```\n\nFor ARM builds, the `COMPILER` flag in `config.mk` needs to changed to `GCCARMv8` or `ARMCLANG` (experimental).\nFor POWER builds, the `COMPILER` flag in `config.mk` needs to changed to `GCCPOWER` or `XLC` (experimental).\nFor Nvidia GPU support, set `NVIDIA_INTERFACE` in `config.mk` to `true` and adjust build-time variables if needed\nFor AMD GPU support, set `ROCM_INTERFACE` in `config.mk` to `true` and adjust build-time variables if needed\n\n--------------------------------------------------------------------------------\nUsage examples\n--------------------------------------------------------------------------------\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-topology\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n--------------------------------------------------------------------------------\nCPU name:\tIntel(R) Core(TM) i7-6700K CPU @ 4.00GHz\nCPU type:\tIntel Skylake processor\nCPU stepping:\t3\n********************************************************************************\nHardware Thread Topology\n********************************************************************************\nSockets:\t\t1\nCores per socket:\t4\nThreads per core:\t2\n--------------------------------------------------------------------------------\nHWThread        Thread        Core        Die        Socket        Available\n0               0             0           0          0             *                \n1               0             1           0          0             *                \n2               0             2           0          0             *                \n3               0             3           0          0             *                \n4               1             0           0          0             *                \n5               1             1           0          0             *                \n6               1             2           0          0             *                \n7               1             3           0          0             *                \n--------------------------------------------------------------------------------\nSocket 0:\t\t( 0 4 1 5 2 6 3 7 )\n--------------------------------------------------------------------------------\n********************************************************************************\nCache Topology\n********************************************************************************\nLevel:\t\t\t1\nSize:\t\t\t32 kB\nCache groups:\t\t( 0 4 ) ( 1 5 ) ( 2 6 ) ( 3 7 )\n--------------------------------------------------------------------------------\nLevel:\t\t\t2\nSize:\t\t\t256 kB\nCache groups:\t\t( 0 4 ) ( 1 5 ) ( 2 6 ) ( 3 7 )\n--------------------------------------------------------------------------------\nLevel:\t\t\t3\nSize:\t\t\t8 MB\nCache groups:\t\t( 0 4 1 5 2 6 3 7 )\n--------------------------------------------------------------------------------\n********************************************************************************\nNUMA Topology\n********************************************************************************\nNUMA domains:\t\t1\n--------------------------------------------------------------------------------\nDomain:\t\t\t0\nProcessors:\t\t( 0 4 1 5 2 6 3 7 )\nDistances:\t\t10\nFree memory:\t\t318.203 MB\nTotal memory:\t\t7626.23 MB\n--------------------------------------------------------------------------------\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-perfctr\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-perfctr -C 0 -g L2 hostname\n--------------------------------------------------------------------------------\nCPU name:\tIntel(R) Core(TM) i7-6700K CPU @ 4.00GHz\nCPU type:\tIntel Skylake processor\nCPU clock:\t4.01 GHz\n--------------------------------------------------------------------------------\nmytesthost\n--------------------------------------------------------------------------------\nGroup 1: L2\n+-----------------------+---------+------------+\n|         Event         | Counter | HWThread 0 |\n+-----------------------+---------+------------+\n|   INSTR_RETIRED_ANY   |  FIXC0  |     321342 |\n| CPU_CLK_UNHALTED_CORE |  FIXC1  |     450498 |\n|  CPU_CLK_UNHALTED_REF |  FIXC2  |    1118900 |\n|    L1D_REPLACEMENT    |   PMC0  |       6670 |\n|      L1D_M_EVICT      |   PMC1  |       1840 |\n| ICACHE_64B_IFTAG_MISS |   PMC2  |       9293 |\n+-----------------------+---------+------------+\n\n+--------------------------------+------------+\n|             Metric             | HWThread 0 |\n+--------------------------------+------------+\n|       Runtime (RDTSC) [s]      |     0.0022 |\n|      Runtime unhalted [s]      |     0.0001 |\n|           Clock [MHz]          |  1613.6392 |\n|               CPI              |     1.4019 |\n|  L2D load bandwidth [MBytes/s] |   197.8326 |\n|  L2D load data volume [GBytes] |     0.0004 |\n| L2D evict bandwidth [MBytes/s] |    54.5745 |\n| L2D evict data volume [GBytes] |     0.0001 |\n|     L2 bandwidth [MBytes/s]    |   528.0381 |\n|     L2 data volume [GBytes]    |     0.0011 |\n+--------------------------------+------------+\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-pin\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-pin -c 0,1,2 ./a.out\n[pthread wrapper] \n[pthread wrapper] MAIN -\u003e 0\n[pthread wrapper] PIN_MASK: 0-\u003e1  1-\u003e2  \n[pthread wrapper] SKIP MASK: 0x0\n\tthreadid 140566548539136 -\u003e hwthread 1 - OK\n\tthreadid 140566540146432 -\u003e hwthread 2 - OK\nNumber of Threads requested = 3\nThread 0 running on processor 0 ....\nThread 1 running on processor 1 ....\nThread 2 running on processor 2 ....\n[...]\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-bench\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-bench -t triad_avx -W N:2GB:3\nWarning: Sanitizing vector length to a multiple of the loop stride 16 and thread count 3 from 62500000 elements (500000000 bytes) to 62499984 elements (499999872 bytes)\nAllocate: Process running on hwthread 0 (Domain N) - Vector length 62499984/499999872 Offset 0 Alignment 512\nAllocate: Process running on hwthread 0 (Domain N) - Vector length 62499984/499999872 Offset 0 Alignment 512\nAllocate: Process running on hwthread 0 (Domain N) - Vector length 62499984/499999872 Offset 0 Alignment 512\nAllocate: Process running on hwthread 0 (Domain N) - Vector length 62499984/499999872 Offset 0 Alignment 512\nInitialization: Each thread in domain initializes its own stream chunks\n--------------------------------------------------------------------------------\nLIKWID MICRO BENCHMARK\nTest: triad_avx\n--------------------------------------------------------------------------------\nUsing 1 work groups\nUsing 3 threads\n--------------------------------------------------------------------------------\nRunning without Marker API. Activate Marker API with -m on commandline.\n--------------------------------------------------------------------------------\nGroup: 0 Thread 1 Global Thread 1 running on hwthread 4 - Vector length 20833328 Offset 20833328\nGroup: 0 Thread 0 Global Thread 0 running on hwthread 0 - Vector length 20833328 Offset 0\nGroup: 0 Thread 2 Global Thread 2 running on hwthread 1 - Vector length 20833328 Offset 41666656\n--------------------------------------------------------------------------------\nCycles:\t\t\t22977763263\nCPU Clock:\t\t4007946861\nCycle Clock:\t\t4007946861\nTime:\t\t\t5.733051e+00 sec\nIterations:\t\t96\nIterations per thread:\t32\nInner loop executions:\t1302083\nSize (Byte):\t\t1999999488\nSize per thread:\t666666496\nNumber of Flops:\t3999998976\nMFlops/s:\t\t697.71\nData volume (Byte):\t63999983616\nMByte/s:\t\t11163.34\nCycles per update:\t11.488885\nCycles per cacheline:\t91.911077\nLoads per update:\t3\nStores per update:\t1\nLoad bytes per element:\t24\nStore bytes per elem.:\t8\nLoad/store ratio:\t3.00\nInstructions:\t\t2374999408\nUOPs:\t\t\t3749999040\n--------------------------------------------------------------------------------\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-mpirun\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-mpirun -mpi slurm -np 4 -t 2 ./a.out\nMPI started\nProcess with rank 0 running on Node f0846.nhr.fau.de core 0\nProcess with rank 2 running on Node f0859.nhr.fau.de core 0\nProcess with rank 3 running on Node f0859.nhr.fau.de core 36\nProcess with rank 1 running on Node f0846.nhr.fau.de core 36\nEnter OpenMP parallel region\nStart OpenMP threads\nRank 0 Thread 0 running on Node f0846.nhr.fau.de core 0\nRank 0 Thread 1 running on Node f0846.nhr.fau.de core 1\nRank 1 Thread 0 running on Node f0846.nhr.fau.de core 36\nRank 1 Thread 1 running on Node f0846.nhr.fau.de core 37\nRank 2 Thread 0 running on Node f0859.nhr.fau.de core 0\nRank 2 Thread 1 running on Node f0859.nhr.fau.de core 1\nRank 3 Thread 0 running on Node f0859.nhr.fau.de core 36\nRank 3 Thread 1 running on Node f0859.nhr.fau.de core 37\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-powermeter\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-powermeter \n--------------------------------------------------------------------------------\nCPU name:\tIntel(R) Core(TM) i7-6700K CPU @ 4.00GHz\nCPU type:\tIntel Skylake processor\nCPU clock:\t4.01 GHz\n--------------------------------------------------------------------------------\n--------------------------------------------------------------------------------\nRuntime: 2.00019 s\nMeasure for socket 0 on CPU 0\nDomain PKG:\nEnergy consumed: 7.47705 Joules\nPower consumed: 3.73817 Watt\nDomain PP0:\nEnergy consumed: 5.42047 Joules\nPower consumed: 2.70998 Watt\nDomain PP1:\nEnergy consumed: 0.0872803 Joules\nPower consumed: 0.043636 Watt\nDomain DRAM:\nEnergy consumed: 1.02612 Joules\nPower consumed: 0.513013 Watt\nDomain PLATFORM:\nEnergy consumed: 0 Joules\nPower consumed: 0 Watt\n--------------------------------------------------------------------------------\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003ccode\u003elikwid-features\u003c/code\u003e\u003c/summary\u003e\n\u003cpre\u003e\n$ likwid-features -c 0 -l\nFeature               HWThread 0\t\nHW_PREFETCHER         on\t\nCL_PREFETCHER         on\t\nDCU_PREFETCHER        on\t\nIP_PREFETCHER         on\t\nFAST_STRINGS          on\t\nTHERMAL_CONTROL       on\t\nPERF_MON              on\t\nFERR_MULTIPLEX        off\t\nBRANCH_TRACE_STORAGE  on\t\nXTPR_MESSAGE          off\t\nPEBS                  on\t\nSPEEDSTEP             on\t\nMONITOR               on\t\nSPEEDSTEP_LOCK        off\t\nCPUID_MAX_VAL         off\t\nXD_BIT                on\t\nDYN_ACCEL             off\t\nTURBO_MODE            on\t\nTM2                   off\n\u003c/pre\u003e\n\u003c/details\u003e\n\n\n--------------------------------------------------------------------------------\nDocumentation\n--------------------------------------------------------------------------------\nFor a detailed  documentation on the usage of the tools have a look at the\nhtml documentation build with doxygen. Call\n\n`make docs`\n\nor after installation, look at the man pages.\n\nThere is also a wiki at the github page:\nhttps://github.com/rrze-likwid/likwid/wiki\n\nIf you have problems or suggestions please let me know on the likwid mailing list:\nhttp://groups.google.com/group/likwid-users\n\nor if it is bug, add an issue at:\nhttps://github.com/rrze-likwid/likwid/issues\n\nYou can also chat with us through Matrix:\n- General chat: https://matrix.to/#/#likwid:matrix.org?via=matrix.org\n- Development chat: https://matrix.to/#/#likwid-dev:matrix.org?via=matrix.org\n\n--------------------------------------------------------------------------------\nExtras\n--------------------------------------------------------------------------------\n- If you want to use the Marker API with Java, you can find the Java module here:\nhttps://github.com/jacek-lewandowski/likwid-java-api\n- For Python you can find an interface to the LIKWID API here:\nhttps://github.com/RRZE-HPC/pylikwid or `pip install pylikwid`\n- A Julia interface to LIKWID is provided by the [Paderborn Center for Parallel Computing (PC²)](https://pc2.uni-paderborn.de) and the [MIT JuliaLab](https://julia.mit.edu/):\nhttps://github.com/JuliaPerf/LIKWID.jl or `] add LIKWID`\n\n--------------------------------------------------------------------------------\nSurvey\n--------------------------------------------------------------------------------\nWe opened a survey at the user mailing list to get a feeling who uses LIKWID and how.\nMoreover we would be interested if you are missing a feature or what annoys you when using LIKWID.\nLink to the survey:\nhttps://groups.google.com/forum/#!topic/likwid-users/F7TDho3k7ps\n\n--------------------------------------------------------------------------------\nFunding\n--------------------------------------------------------------------------------\n\nLIKWID development was funded by BMBF Germany under the [FEPA project](https://gauss-allianz.de/en/project/title/FEPA), grant 01IH13009. Since 2017 the development is further funded by BMBF Germany under the [SeASiTe project](https://gauss-allianz.de/en/project/title/SeASiTe), grant 01IH16012A. In 2022, the [EE-HPC project](https://gauss-allianz.de/en/project/title/EE-HPC) is funded by BMBF Germany in the GreenHPC grant.\n\n\u003cdiv align=center\u003e\u003cimg src=\"https://raw.githubusercontent.com/wiki/RRZE-HPC/likwid/images/BMBF.png\" alt=\"BMBF logo\" width=\"150\"/\u003e\u003c/div\u003e\n","funding_links":[],"categories":["C","Software","Debugging and Profiling Tools 🔍","1. System Overview","Nix tools"],"sub_categories":["Trends","Language-Specific Libraries 🔤"],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FRRZE-HPC%2Flikwid","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FRRZE-HPC%2Flikwid","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FRRZE-HPC%2Flikwid/lists"}