Skip to content
benchx
Template
Search
benchx-dev/benchx
Home
Design
Schemas
Use cases
User stories
benchx
Releasing
benchx
benchx-dev/benchx
Home
Design
Design
Overview
Benchmark result schema
System decomposition
Runner
Runner schemas
Harness adapter
Benchmark environments
Comparator
Prototype scope
Prototype design
Schemas
Schemas
Overview
Measurement result
Work order
Target description
Target provider request
Environment policy
Environment identity
Identity policy
Comparison document
Project configuration
Use cases
Use cases
Overview
Template
Template
On this page
UC-NN:
</span> </span> </a> <nav class="md-nav" aria-label="UC-NN:"> <ul class="md-nav__list"> <li class="md-nav__item"> <a href="#1-summary" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 1. Summary </span> </span> </a> </li> <li class="md-nav__item"> <a href="#2-motivation" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 2. Motivation </span> </span> </a> </li> <li class="md-nav__item"> <a href="#3-actors-and-trigger" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 3. Actors and trigger </span> </span> </a> </li> <li class="md-nav__item"> <a href="#4-workflow-sketch" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 4. Workflow sketch </span> </span> </a> </li> <li class="md-nav__item"> <a href="#5-study-design" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 5. Study design </span> </span> </a> </li> <li class="md-nav__item"> <a href="#6-measurements" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 6. Measurements </span> </span> </a> </li> <li class="md-nav__item"> <a href="#7-comparison-semantics" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 7. Comparison semantics </span> </span> </a> </li> <li class="md-nav__item"> <a href="#8-schema-implications" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 8. Schema implications </span> </span> </a> </li> <li class="md-nav__item"> <a href="#9-storage-and-lifecycle" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 9. Storage and lifecycle </span> </span> </a> </li> <li class="md-nav__item"> <a href="#10-degenerate-and-failure-modes" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 10. Degenerate and failure modes </span> </span> </a> </li> <li class="md-nav__item"> <a href="#11-non-goals" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 11. Non-goals </span> </span> </a> </li> <li class="md-nav__item"> <a href="#12-open-questions" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 12. Open questions </span> </span> </a> </li> </ul> </nav> </li> <li class="md-nav__item"> <a href="#worked-example" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> Worked example </span> </span> </a> <nav class="md-nav" aria-label="Worked example"> <ul class="md-nav__list"> <li class="md-nav__item"> <a href="#uc-03-compare-one-benchmark-across-contributors-machines" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> UC-03: Compare one benchmark across contributors' machines </span> </span> </a> </li> </ul> </nav> </li> </ul> </nav> </li> <li class="md-nav__item"> <a href="../UC-01-short-slug/" class="md-nav__link"> <span class="md-ellipsis"> UC-01 Single parametrized benchmark </span> </a> </li> <li class="md-nav__item"> <a href="../UC-02-variants/" class="md-nav__link"> <span class="md-ellipsis"> UC-02 Variants </span> </a> </li> <li class="md-nav__item"> <a href="../UC-03-compare/" class="md-nav__link"> <span class="md-ellipsis"> UC-03 Compare </span> </a> </li> <li class="md-nav__item"> <a href="../UC-04-cross-env/" class="md-nav__link"> <span class="md-ellipsis"> UC-04 Cross-environment </span> </a> </li> </ul> </nav> </li> <li class="md-nav__item md-nav__item--nested"> <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_5" > <label class="md-nav__link" for="__nav_5" id="__nav_5_label" tabindex="0"> <span class="md-ellipsis"> User stories </span> <span class="md-nav__icon md-icon"></span> </label> <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_5_label" aria-expanded="false"> <label class="md-nav__title" for="__nav_5"> <span class="md-nav__icon md-icon"></span> User stories </label> <ul class="md-nav__list" data-md-scrollfix> <li class="md-nav__item"> <a href="../../user_stories/" class="md-nav__link"> <span class="md-ellipsis"> Overview </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/TEMPLATE/" class="md-nav__link"> <span class="md-ellipsis"> Template </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/apache-arrow/" class="md-nav__link"> <span class="md-ellipsis"> Apache Arrow </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/apache-arrow-local-benchmarking/" class="md-nav__link"> <span class="md-ellipsis"> Apache Arrow (local) </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/numpy/" class="md-nav__link"> <span class="md-ellipsis"> NumPy </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/pandas/" class="md-nav__link"> <span class="md-ellipsis"> pandas </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/narwhals/" class="md-nav__link"> <span class="md-ellipsis"> Narwhals </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/openblas_user_stories/" class="md-nav__link"> <span class="md-ellipsis"> OpenBLAS </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/array-api-and-accelerators/" class="md-nav__link"> <span class="md-ellipsis"> Array API and accelerators </span> </a> </li> <li class="md-nav__item"> <a href="../../user_stories/one-off/" class="md-nav__link"> <span class="md-ellipsis"> One-off benchmarking </span> </a> </li> </ul> </nav> </li> <li class="md-nav__item"> <a href="../../benchx/" class="md-nav__link"> <span class="md-ellipsis"> benchx </span> </a> </li> <li class="md-nav__item"> <a href="../../releasing/" class="md-nav__link"> <span class="md-ellipsis"> Releasing </span> </a> </li> </ul> </nav> </div> </div> </div> <div class="md-sidebar md-sidebar--secondary" data-md-component="sidebar" data-md-type="toc" > <div class="md-sidebar__scrollwrap"> <input class="md-nav__toggle md-toggle" type="checkbox" id="__toc"> <div class="md-sidebar-button__wrapper"> <label class="md-sidebar-button" for="__toc"></label> </div> <div class="md-sidebar__inner"> <nav class="md-nav md-nav--secondary" aria-label="On this page"> <label class="md-nav__title" for="__toc"> <span class="md-nav__icon md-icon"></span> On this page </label> <ul class="md-nav__list" data-md-component="toc" data-md-scrollfix> <li class="md-nav__item"> <a href="#uc-nn" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> UC-NN: <title> </span> </span> </a> <nav class="md-nav" aria-label="UC-NN:"> <ul class="md-nav__list"> <li class="md-nav__item"> <a href="#1-summary" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 1. Summary </span> </span> </a> </li> <li class="md-nav__item"> <a href="#2-motivation" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 2. Motivation </span> </span> </a> </li> <li class="md-nav__item"> <a href="#3-actors-and-trigger" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 3. Actors and trigger </span> </span> </a> </li> <li class="md-nav__item"> <a href="#4-workflow-sketch" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 4. Workflow sketch </span> </span> </a> </li> <li class="md-nav__item"> <a href="#5-study-design" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 5. Study design </span> </span> </a> </li> <li class="md-nav__item"> <a href="#6-measurements" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 6. Measurements </span> </span> </a> </li> <li class="md-nav__item"> <a href="#7-comparison-semantics" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 7. Comparison semantics </span> </span> </a> </li> <li class="md-nav__item"> <a href="#8-schema-implications" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 8. Schema implications </span> </span> </a> </li> <li class="md-nav__item"> <a href="#9-storage-and-lifecycle" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 9. Storage and lifecycle </span> </span> </a> </li> <li class="md-nav__item"> <a href="#10-degenerate-and-failure-modes" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 10. Degenerate and failure modes </span> </span> </a> </li> <li class="md-nav__item"> <a href="#11-non-goals" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 11. Non-goals </span> </span> </a> </li> <li class="md-nav__item"> <a href="#12-open-questions" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> 12. Open questions </span> </span> </a> </li> </ul> </nav> </li> <li class="md-nav__item"> <a href="#worked-example" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> Worked example </span> </span> </a> <nav class="md-nav" aria-label="Worked example"> <ul class="md-nav__list"> <li class="md-nav__item"> <a href="#uc-03-compare-one-benchmark-across-contributors-machines" class="md-nav__link"> <span class="md-ellipsis"> <span class="md-typeset"> UC-03: Compare one benchmark across contributors' machines </span> </span> </a> </li> </ul> </nav> </li> </ul> </nav> </div> </div> </div> <div class="md-content" data-md-component="content"> <article class="md-content__inner md-typeset"> <a id="__skip" tabindex="-1"></a> <a href="https://github.com/benchx-dev/benchx/edit/main/docs/use-cases/UC_NN_TEMPLATE.md" title="Edit this page" class="md-content__button md-icon" rel="edit"> <svg xmlns="http://www.w3.org/2000/svg" fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" class="lucide lucide-file-pen" viewBox="0 0 24 24"><path d="M12.659 22H18a2 2 0 0 0 2-2V8a2.4 2.4 0 0 0-.706-1.706l-3.588-3.588A2.4 2.4 0 0 0 14 2H6a2 2 0 0 0-2 2v9.34"/><path d="M14 2v5a1 1 0 0 0 1 1h5M10.378 12.622a1 1 0 0 1 3 3.003L8.36 20.637a2 2 0 0 1-.854.506l-2.867.837a.5.5 0 0 1-.62-.62l.836-2.869a2 2 0 0 1 .506-.853z"/></svg> </a> <h1 id="benchmark-use-case-template">Benchmark Use Case Template<a class="headerlink" href="#benchmark-use-case-template" title="Permanent link">¶</a></h1> <p>One file per use case, in <code>docs/use-cases/</code>, named <code>UC-NN-short-slug.md</code>.</p> <p><strong>How to use this.</strong> Fill sections 1-6 from the user's perspective; they should be readable by someone who has never seen the schema. Sections 7-10 are the payoff - they are what the schema is derived from. A use case that cannot complete section 8 is not yet understood well enough to design against.</p> <p>Delete guidance text as you go. Write "n/a" rather than deleting a heading, so the shape stays comparable across documents.</p> <hr /> <h2 id="uc-nn">UC-NN: <title><a class="headerlink" href="#uc-nn" title="Permanent link">¶</a></h2> <table> <thead> <tr> <th></th> <th></th> </tr> </thead> <tbody> <tr> <td><strong>Status</strong></td> <td>draft / under review / accepted / rejected / superseded by UC-NN</td> </tr> <tr> <td><strong>Owner</strong></td> <td></td> </tr> <tr> <td><strong>Last updated</strong></td> <td></td> </tr> <tr> <td><strong>Related</strong></td> <td>UC-NN, UC-NN</td> </tr> </tbody> </table> <h3 id="1-summary">1. Summary<a class="headerlink" href="#1-summary" title="Permanent link">¶</a></h3> <p>One or two sentences. What does someone accomplish, and why can't they do it today?</p> <h3 id="2-motivation">2. Motivation<a class="headerlink" href="#2-motivation" title="Permanent link">¶</a></h3> <p>Who has this problem and how often? What do they currently do instead, and what does that cost them - wrong conclusions, wasted time, regressions shipped? Prefer a real incident or a named workflow over a hypothetical.</p> <p>If this use case exists mainly because a tool we might adopt happens to support it, say so. That is a weaker justification and should be visible as such.</p> <h3 id="3-actors-and-trigger">3. Actors and trigger<a class="headerlink" href="#3-actors-and-trigger" title="Permanent link">¶</a></h3> <ul> <li><strong>Actor:</strong> contributor on a laptop / maintainer reviewing a PR / CI on merge / release manager / downstream packager / someone else</li> <li><strong>Trigger:</strong> what makes them start? A command they type, a CI event, a schedule, an alert.</li> <li><strong>Frequency:</strong> per keystroke-loop / per PR / per merge / nightly / per release</li> </ul> <h3 id="4-workflow-sketch">4. Workflow sketch<a class="headerlink" href="#4-workflow-sketch" title="Permanent link">¶</a></h3> <p>The commands or UI actions, in order, including what the actor sees back. Rough is fine; the point is to expose where data is produced and consumed.</p> <div class="highlight"><pre><span></span><code>$ <command> <what they see> </code></pre></div> <p>Note where the loop closes: what do they do with the answer? If the answer is "nothing, they look at it," say that - some use cases really are just display, and those have much weaker schema requirements than ones feeding an automated decision.</p> <h3 id="5-study-design">5. Study design<a class="headerlink" href="#5-study-design" title="Permanent link">¶</a></h3> <p>The core of the use case. Every record is a measurement plus coordinates:</p> <ul> <li><strong>Code identity</strong> - benchmark ID, parameters</li> <li><strong>Code version</strong> - commit, patch/variant label, build configuration</li> <li><strong>Environment</strong> - machine, OS, CPU, BLAS vendor and threading, tuning state</li> <li><strong>Execution context</strong> - session, round, interleave position, instrument</li> </ul> <p>State for each:</p> <table> <thead> <tr> <th>Coordinate</th> <th>Role</th> <th>Notes</th> </tr> </thead> <tbody> <tr> <td>Code identity</td> <td>varying / controlled / free</td> <td></td> </tr> <tr> <td>Code version</td> <td>varying / controlled / free</td> <td></td> </tr> <tr> <td>Environment</td> <td>varying / controlled / free</td> <td></td> </tr> <tr> <td>Execution context</td> <td>varying / controlled / free</td> <td></td> </tr> </tbody> </table> <ul> <li><strong>Varying</strong> - the independent variable; identifying fields must be present and <em>distinct</em> across records being compared.</li> <li><strong>Controlled</strong> - must be present and <em>equal</em> across records being compared.</li> <li><strong>Free</strong> - genuinely irrelevant to the comparison's validity. Justify each one; "free" is where invalid comparisons hide.</li> </ul> <p>If more than one coordinate varies, explain why the comparison is still meaningful, or split this into multiple use cases.</p> <h3 id="6-measurements">6. Measurements<a class="headerlink" href="#6-measurements" title="Permanent link">¶</a></h3> <table> <thead> <tr> <th>Measure</th> <th>Instrument</th> <th>Unit</th> <th>Direction</th> </tr> </thead> <tbody> <tr> <td>e.g. wall time</td> <td>pyperf</td> <td>s</td> <td>lower is better</td> </tr> <tr> <td>e.g. peak RSS</td> <td>memray</td> <td>bytes</td> <td>lower is better</td> </tr> </tbody> </table> <ul> <li><strong>Raw samples retained?</strong> yes / no / how many</li> <li><strong>Is the instrument available in every environment this use case spans?</strong> (Cross-platform and cross-machine cases frequently fail here.)</li> </ul> <h3 id="7-comparison-semantics">7. Comparison semantics<a class="headerlink" href="#7-comparison-semantics" title="Permanent link">¶</a></h3> <ul> <li><strong>What comparison is valid?</strong> paired / unpaired / against a stored baseline / against a machine-relative anchor / none, absolute values only</li> <li><strong>Estimator:</strong> median, trimmed mean, minimum, full distribution</li> <li><strong>What counts as a real difference?</strong> A percentage, a significance criterion, or "human eyeballs it." Be concrete - this determines whether raw samples and environment strictness are load-bearing or decorative.</li> <li><strong>Expected noise floor</strong> in this setting, and what it implies about the smallest detectable effect.</li> </ul> <h3 id="8-schema-implications">8. Schema implications<a class="headerlink" href="#8-schema-implications" title="Permanent link">¶</a></h3> <p><strong>The section this document exists for.</strong></p> <ul> <li><strong>Profile:</strong> name of the validation profile this needs. Existing one, or new?</li> <li><strong>Fields required beyond the core:</strong></li> <li><strong>Fields required to be <em>absent</em> or explicitly null:</strong></li> <li><strong>New fields not currently in the schema:</strong> name, type, intended Arrow type, who populates it, and what happens when it can't be determined.</li> <li><strong>Comparability key:</strong> which environment fields must match for two records in this use case to belong to the same series?</li> <li><strong>Validation invariants:</strong> e.g. "all records in a comparison share <code>env_hash</code>"; "<code>session_id</code> equal and <code>round</code> distinct"; "<code>anchor_run_id</code> present and resolvable".</li> <li><strong>Conflicts with existing use cases:</strong> a field this needs required that another needs optional, a different meaning for a shared field, an incompatible key.</li> </ul> <h3 id="9-storage-and-lifecycle">9. Storage and lifecycle<a class="headerlink" href="#9-storage-and-lifecycle" title="Permanent link">¶</a></h3> <ul> <li><strong>Volume:</strong> records per invocation, invocations per day, per actor</li> <li><strong>Retention:</strong> ephemeral (discarded after the loop) / session-scoped / permanent</li> <li><strong>Location:</strong> local working dir / shared store / both</li> <li><strong>Does this data ever need to join against data from another use case?</strong> If yes, name it - this is the main argument for a shared core, and the main source of pressure on it.</li> </ul> <h3 id="10-degenerate-and-failure-modes">10. Degenerate and failure modes<a class="headerlink" href="#10-degenerate-and-failure-modes" title="Permanent link">¶</a></h3> <ul> <li>What happens when a <em>controlled</em> coordinate silently isn't constant? (Machine thermal state drifts; BLAS changes under an unpinned environment; the working tree is dirty.) Is that detectable from the record alone, or does bad data enter the store looking valid?</li> <li>What happens with missing or partial data - one variant failed, a machine dropped out, an instrument is unavailable?</li> <li>What is the worst wrong conclusion someone could draw, and does the schema make it harder or easier to draw?</li> </ul> <h3 id="11-non-goals">11. Non-goals<a class="headerlink" href="#11-non-goals" title="Permanent link">¶</a></h3> <p>What this use case explicitly does <em>not</em> need to support. Especially: adjacent things a reader would reasonably assume are included.</p> <h3 id="12-open-questions">12. Open questions<a class="headerlink" href="#12-open-questions" title="Permanent link">¶</a></h3> <p>Numbered, so they can be referenced in review.</p> <hr /> <h2 id="worked-example">Worked example<a class="headerlink" href="#worked-example" title="Permanent link">¶</a></h2> <h3 id="uc-03-compare-one-benchmark-across-contributors-machines">UC-03: Compare one benchmark across contributors' machines<a class="headerlink" href="#uc-03-compare-one-benchmark-across-contributors-machines" title="Permanent link">¶</a></h3> <table> <thead> <tr> <th></th> <th></th> </tr> </thead> <tbody> <tr> <td><strong>Status</strong></td> <td>draft</td> </tr> </tbody> </table> <p><strong>1. Summary.</strong> A contributor sees a slowdown that another contributor cannot reproduce. They want to run the same benchmark, at the same commit, on both machines and determine whether the difference is real and environment-dependent.</p> <p><strong>5. Study design.</strong></p> <table> <thead> <tr> <th>Coordinate</th> <th>Role</th> <th>Notes</th> </tr> </thead> <tbody> <tr> <td>Code identity</td> <td>controlled</td> <td>same benchmark ID and params</td> </tr> <tr> <td>Code version</td> <td>controlled</td> <td>same commit; both trees must be clean</td> </tr> <tr> <td>Environment</td> <td><strong>varying</strong></td> <td>the independent variable</td> </tr> <tr> <td>Execution context</td> <td>free</td> <td>sessions are inherently separate</td> </tr> </tbody> </table> <p><strong>7. Comparison semantics.</strong> Absolute wall time is not comparable across machines. Comparison must be against a per-machine anchor workload run in the same session. Meaningful difference: a ratio-of-ratios, not a raw delta.</p> <p><strong>8. Schema implications.</strong></p> <ul> <li>Profile: new, <code>cross-environment</code>.</li> <li>Environment fields become <em>mandatory and maximal</em> - CPU model, microcode where available, governor, turbo state, BLAS vendor and version, thread caps, build flags. This is the strictest environment requirement of any use case so far, despite being a purely local workflow.</li> <li>New field: <code>anchor_run_id</code>, referencing an anchor benchmark record from the same session on the same machine. Nullable only if the profile is not claimed.</li> <li>Comparability key: inverted - records must <em>differ</em> in machine identity and <em>agree</em> on everything under code identity and code version.</li> <li>Conflict: contradicts the assumption that local records may omit build metadata.</li> </ul> <p><strong>10. Degenerate modes.</strong> If one contributor's tree is dirty, the commit field is a lie and the comparison is meaningless while still appearing valid. The schema should make <code>dirty</code> non-nullable in this profile so the record cannot be produced silently.</p> </article> </div> <script>var target=document.getElementById(location.hash.slice(1));target&&target.name&&(target.checked=target.name.startsWith("__tabbed_"))</script> </div> <button type="button" class="md-top md-icon" data-md-component="top" hidden> <svg xmlns="http://www.w3.org/2000/svg" fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" class="lucide lucide-circle-arrow-up" viewBox="0 0 24 24"><circle cx="12" cy="12" r="10"/><path d="m16 12-4-4-4 4M12 16V8"/></svg> Back to top </button> </main> <footer class="md-footer"> <nav class="md-footer__inner md-grid" aria-label="Footer" > <a href="../" class="md-footer__link md-footer__link--prev" aria-label="Previous: Overview"> <div class="md-footer__button md-icon"> <svg xmlns="http://www.w3.org/2000/svg" fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" class="lucide lucide-arrow-left" viewBox="0 0 24 24"><path d="m12 19-7-7 7-7M19 12H5"/></svg> </div> <div class="md-footer__title"> <span class="md-footer__direction"> Previous </span> <div class="md-ellipsis"> Overview </div> </div> </a> <a href="../UC-01-short-slug/" class="md-footer__link md-footer__link--next" aria-label="Next: UC-01 Single parametrized benchmark"> <div class="md-footer__title"> <span class="md-footer__direction"> Next </span> <div class="md-ellipsis"> UC-01 Single parametrized benchmark </div> </div> <div class="md-footer__button md-icon"> <svg xmlns="http://www.w3.org/2000/svg" fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" class="lucide lucide-arrow-right" viewBox="0 0 24 24"><path d="M5 12h14M12 5l7 7-7 7"/></svg> </div> </a> </nav> <div class="md-footer-meta md-typeset"> <div class="md-footer-meta__inner md-grid"> <div class="md-copyright"> <div class="md-copyright__highlight"> Copyright © benchx contributors </div> Made with <a href="https://zensical.org/" target="_blank" rel="noopener"> Zensical </a> </div> </div> </div> </footer> </div> <div class="md-dialog" data-md-component="dialog" role="status" aria-atomic="true"> <div class="md-dialog__inner md-typeset"></div> </div> <script id="__config" type="application/json">{"annotate":null,"base":"../..","features":["content.action.edit","content.code.copy","navigation.footer","navigation.sections","navigation.tabs","navigation.tabs.sticky","navigation.top","search.highlight","search.suggest","toc.follow"],"redirect":null,"search":"../../assets/javascripts/workers/search.c47bb847.min.js","tags":null,"translations":{"clipboard.copied":"Copied to clipboard","clipboard.copy":"Copy to clipboard","clipboard.error":"Could not copy to clipboard","search.result.more.one":"1 more on this page","search.result.more.other":"# more on this page","search.result.none":"No matching documents","search.result.one":"1 matching document","search.result.other":"# matching documents","search.result.placeholder":"Type to start searching","search.result.term.missing":"Missing","select.version":"Select version"},"version":null}</script> <script src="../../assets/javascripts/bundle.a2388154.min.js"></script> </body> </html>