Meta tags:
description= CUDA Kernel Benchmarking Library. Contribute to NVIDIA/nvbench development by creating an account on GitHub.;
Headings (most frequently used words):
navigation, and, repository, about, files, examples, footer, topics, resources, code, of, conduct, stars, watchers, forks, overview, supported, compilers, tools, getting, started, contributing, license, scope, related, projects, menu, latest, commit, history, folders, minimal, benchmark, command, line, interface, demo, project, tests, releases, used, by, contributors, languages, building,
Text of the page (most frequently used words):
and (31), #nvbench (26), for (19), the (18), benchmark (17), github (15), cuda (15), gpu (13), build (12), time (11), code (11), with (11), cpu (10), kernel (9), are (9), you (8), license (8), benchmarking (8), repository (7), examples (7), parameter (6), more (6), cmake (6), this (5), security (5), conduct (5), readme (5), nvidia (5), kernels (5), library (5), see (5), all (5), example (5), your (5), throughput (5), total (5), output (5), features (5), commit (5), that (4), navigation (4), forks (4), 137 (4), stars (4), 949 (4), apache (4), performance (4), provides (4), benchmarks (4), execution (4), multiple (4), tests (4), default (4), project (4), out (4), command (4), line (4), memory (4), device (4), batch (4), txt (4), agents (4), yaml (4), signed (4), support (4), copilot (4), not (3), information (3), docs (3), report (3), resources (3), about (3), only (3), change (3), testing (3), projects (3), make (3), bin (3), issues (3), number (3), global (3), mib (3), sec (3), kib (3), block (3), noise (3), runtime (3), sweeps (3), overview (3), state (3), launch (3), exec (3), clang (3), measurements (3), files (3), review (3), actions (3), quality (3), another (3), tab (3), window (3), refresh (3), session (3), reload (3), sign (3), developers (3), changelog (3), customer (3), enterprise (3), grade (3), can (2), community (2), status (2), footer (2), contributors (2), watching (2), watchers (2), custom (2), properties (2), activity (2), topics (2), intended (2), non (2), but (2), may (2), consider (2), using (2), per (2), analysis (2), end (2), tools (2), host (2), side (2), test (2), run (2), built (2), into (2), prefixed (2), mkdir (2), current (2), started (2), demo (2), devices (2), quadro (2), gv100 (2), version (2), 700 (2), peak (2), gib (2), cache (2), maximum (2), active (2), 65536 (2), throughput_bench (2), measurement (2), pass (2), cold (2), 27393x (2), 27394x (2), set (2), manual (2), timing (2), exec_tag (2), timer (2), sync (2), markdown (2), item (2), compile (2), axes (2), each (2), rich (2), cli (2), axis (2), interface (2), my_benchmark (2), types (2), times (2), back (2), formats (2), second (2), work (2), items (2), requirements (2), pyproject (2), toml (2), code_of_conduct (2), cmakepresets (2), json (2), cmakelists (2), pre (2), config (2), gitignore (2), git (2), blame (2), ignore (2), revs (2), coderabbit (2), clangd (2), format (2), python (2), devcontainer (2), skills (2), last (2), message (2), name (2), 840 (2), commits (2), latest (2), menu (2), must (2), star (2), settings (2), insights (2), universe (2), training (2), integrations (2), stories (2), center (2), documentation (2), learn (2), business (2), fixes (2), products (2), secret (2), perform, action, share, personal, manage, cookies, contact, privacy, terms, 2026, inc, languages, used, releases, focused, evaluating, also, facilities, trivial, workloads, optimized, microbenchmarks, future, now, google, high, resolution, will, measure, regression, tuning, individual, depth, applications, nsight, appropriate, single, critical, region, scope, related, released, under, llvm, exceptions, ctest, dnvbench_enable_testing, labeled, good, first, issue, board, contributions, welcome, contributing, simple, uses, great, way, experiment, without, lot, investment, nvbench_demo, get, own, trying, ptx, sms, clock, rate, 1627, mhz, 32163, free, 32508, bus, 870, 4096, bit, ddr, 850mhz, max, shared, size, 6144, blocks, threads, 2048, 1024, available, registers, ecc, enabled, log, warn, timed, 00s, while, over, threshold, 262392ms, 267860ms, 19s, 261963ms, 18s, results, numelements, datasize, samples, elem, globalmem, bwpeak, 16777216, 000, 267, 860, 262, 392, 940g, 476, 387, 261, 963, from, sure, based, running, cmake_cuda_architecture, dnvbench_enable_examples, dcmake_cuda_architectures, building, adding, hiding, columns, summaries, specific, stream, skipping, configurations, reporting, byte, statistics, enums, constant, integral, demonstrate, various, usecases, executable, produced, options, configuring, specification, customizing, implementing, my_kernel, num_blocks, nvbench_bench, get_stream, 256, void, basic, created, just, few, lines, minimal, getting, headers, tested, toolkit, nvcc, above, supported, compilers, check, challenges, inherent, how, solves, them, talk, suitable, microbenchmarking, measures, reports, average, executions, executes, records, reported, sample, runs, once, clean, optional, explicitly, start, stop, implementation, mode, currently, supports, csv, compute, bandwidth, usage, bytes, elements, calculations, allows, selection, locking, clocks, volta, changing, redefinition, customization, powerful, flexible, system, explores, configuration, space, parameters, dynamic, numbers, strings, static, designed, simplify, progress, everything, subject, view, date, folders, history, open, file, tags, branches, main, fork, notification, notifications, public, discussions, pull, requests, dismiss, alert, switched, accounts, appearance, search, pricing, event, talks, workshops, what, next, partners, company, forum, events, webinars, blog, resource, connect, premium, advanced, add, ons, trust, why, evaluate, industries, government, manufacturing, financial, services, healthcare, industry, nonprofits, startups, team, enterprises, stage, businesses, source, improvements, across, one, place, developer, articles, sponsors, trending, repos, explore, customize, certifications, scanning, leak, prevention, protection, vulnerability, detection, remediation, secure, cloud, development, environments, codespaces, enforcement, automation, ship, coding, agent, tasks, automated, hosting, collaboration, platform, skip, content,
Text of the page (random words):
github nvidia nvbench cuda kernel benchmarking library github skip to content navigation menu sign in products platform github repository hosting code review and collaboration copilot ai coding agent for tasks projects and automated work build ship actions ci cd and automation code quality quality analysis fixes and enforcement codespaces cloud development environments secure code security vulnerability detection and remediation secret protection secret scanning and leak prevention all features developers learn documentation training certifications build build with copilot customize copilot integrations explore trending repos github sponsors developer articles changelog changelog your source for the latest features improvements and fixes across github products all in one place businesses by stage enterprises team startups nonprofits by industry healthcare financial services manufacturing government see all industries evaluate github why github customer stories business insights trust center add ons github advanced security enterprise grade code security copilot for business enterprise grade ai features premium support enterprise grade 24 7 support resources learn connect documentation resource center blog events webinars community forum customer stories integrations support customer support training github status company about partners changelog github universe github universe the event for developers by developers talks workshops and what s next for github pricing search sign in sign up appearance settings you signed in with another tab or window reload to refresh your session you signed out in another tab or window reload to refresh your session you switched accounts on another tab or window reload to refresh your session dismiss alert message nvidia nvbench repository navigation code issues 63 63 pull requests 23 23 discussions actions projects security and quality insights more items nvbench public notifications you must be signed in to change notification settings fork 137 137 star 949 949 you must be signed in to star a repository about cuda kernel benchmarking library topics benchmark cuda cuda kernels gpu kernel benchmark nvidia performance resources readme apache 2 0 license code of conduct code of conduct activity custom properties stars 949 stars watchers 15 watching forks 137 forks report repository main branches tags go to file code open more actions menu latest commit history 840 commits 840 commits folders and files name name last commit message last commit date agents skills review nvbench agents skills review nvbench devcontainer devcontainer github github ci ci cmake cmake docs docs examples examples exec exec nvbench nvbench python python testing testing clang format clang format clangd clangd coderabbit yaml coderabbit yaml git blame ignore revs git blame ignore revs gitignore gitignore pre commit config yaml pre commit config yaml agents md agents md cmakelists txt cmakelists txt cmakepresets json cmakepresets json code_of_conduct md code_of_conduct md license license readme md readme md pyproject toml pyproject toml requirements txt requirements txt view all files repository files navigation readme code of conduct apache 2 0 license more items overview this project is a work in progress everything is subject to change nvbench is a c 17 library designed to simplify cuda kernel benchmarking it features parameter sweeps a powerful and flexible axis system explores a kernel s configuration space parameters may be dynamic numbers strings or static types runtime customization a rich command line interface allows redefinition of parameter axes cuda device selection locking gpu clocks volta changing output formats and more throughput calculations compute and report item throughput elements second global memory bandwidth usage bytes second and per device of peak bw multiple output formats currently supports markdown default and csv output manual timer mode optional explicitly start stop timing in a benchmark implementation multiple measurement types cold measurements each sample runs the benchmark once with a clean device l2 cache gpu and cpu times are reported batch measurements executes the benchmark multiple times back to back and records total time reports the average execution time total time number of executions cpu only measurements measures the host side execution time of a non gpu benchmark not suitable for microbenchmarking check out this talk for an overview of the challenges inherent to cuda kernel benchmarking and how nvbench solves them for you supported compilers and tools cmake 3 30 4 cuda toolkit nvcc 12 0 and above g 7 14 clang 14 19 headers are tested with c 17 c 20 getting started minimal benchmark a basic kernel benchmark can be created with just a few lines of cuda c void my_benchmark nvbench state state state exec nvbench launch launch my_kernel num_blocks 256 0 launch get_stream nvbench_bench my_benchmark see benchmarks for information on customizing benchmarks and implementing parameter sweeps command line interface each benchmark executable produced by nvbench provides a rich set of command line options for configuring benchmark execution at runtime see the cli overview and cli axis specification for more information examples this repository provides a number of examples that demonstrate various nvbench features and usecases runtime and compile time parameter sweeps cpu only benchmarking enums and compile time constant integral parameter axes reporting item sec and byte sec throughput statistics skipping benchmark configurations benchmarking on a specific stream adding hiding columns summaries in markdown output benchmarks that sync cuda devices nvbench exec_tag sync manual timing nvbench exec_tag timer building examples to build the examples mkdir p build cd build cmake dnvbench_enable_examples on dcmake_cuda_architectures 70 make be sure to set cmake_cuda_architecture based on the gpu you are running on examples are built by default into build bin and are prefixed with nvbench example example output from nvbench example throughput command line bin nvbench example throughput devices 0 quadro gv100 sm version 700 ptx version 700 number of sms 80 sm default clock rate 1627 mhz global memory 32163 mib free 32508 mib total global memory bus peak 870 gib sec 4096 bit ddr 850mhz max shared memory 96 kib sm 48 kib block l2 cache size 6144 kib maximum active blocks 32 sm maximum active threads 2048 sm 1024 block available registers 65536 sm 65536 block ecc enabled no log run throughput_bench device 0 warn current measurement timed out 15 00s while over noise threshold 1 26 0 50 pass cold 0 262392ms gpu 0 267860ms cpu 7 19s total gpu 27393x pass batch 0 261963ms gpu 7 18s total gpu 27394x benchmark results throughput_bench 0 quadro gv100 numelements datasize samples cpu time noise gpu time noise elem s globalmem bw bwpeak batch gpu batch 16777216 64 000 mib 27393x 267 860 us 1 25 262 392 us 1 26 63 940g 476 387 gib s 58 77 261 963 us 27394x demo project to get started using nvbench with your own kernels consider trying out the nvbench demo project nvbench_demo provides a simple cmake project that uses nvbench to build an example benchmark it s a great way to experiment with the library without a lot of investment contributing contributions are welcome for current issues see the issue board issues labeled with are good for first time contributors tests to build nvbench tests mkdir p build cd build cmake dnvbench_enable_testing on make tests are built by default into build bin and prefixed with nvbench test to run all tests make test or ctest license nvbench is released under the apache 2 0 license with llvm exceptions see license scope and related projects nvbench will measure the cpu and cuda gpu execution time of a single host side critical region per benchmark it is intended for regression testing and parameter tuning of individual kernels for in depth analysis of end to end performance of multiple applications the nvidia nsight tools are more appropriate nvbench is focused on evaluating the performance of cuda kernels it also provides cpu only benchmarking facilities intended for non trivial cpu workloads but is not optimized for cpu microbenchmarks this may change in the future but for now consider using google benchmark for high resolution cpu benchmarks about cuda kernel benchmarking library topics benchmark cuda cuda kernels gpu kernel benchmark nvidia performance resources readme apache 2 0 license code of conduct code of conduct activity custom properties stars 949 stars watchers 15 watching forks 137 forks report repository releases used by contributors languages footer 2026 github inc footer navigation terms privacy security status community docs contact manage cookies do not share my personal information you can t perform that action at this time
|