--- title: Optimizing Arm-based Build Servers | daily.dev description: A practical guide to optimizing Linux build servers on Ampere Arm-based hardware (Altra and AmpereOne). Covers enabling parallel builds across all cores, using... canonical: https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Optimizing Arm-based Build Servers | daily.dev og:description: A practical guide to optimizing Linux build servers on Ampere Arm-based hardware (Altra and AmpereOne). Covers enabling parallel builds across all cores, using... og:url: https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm og:image: https://api.daily.dev/og/posts/Rh07W05Dm.png og:image:alt: Optimizing Arm-based Build Servers og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Optimizing Arm-based Build Servers **[SitePoint](https://daily.dev/sources/sitepoint)** · 14 min read · 1 upvotes · 0 comments ## Summary A practical guide to optimizing Linux build servers on Ampere Arm-based hardware (Altra and AmpereOne). Covers enabling parallel builds across all cores, using RAM disks (tmpfs) to eliminate disk I/O bottlenecks, setting the CPU frequency governor to performance mode (yielding a 17% speedup), and tuning the kernel CONFIG_HZ parameter to 100 Hz to reduce context-switch overhead (2% additional gain). The LLVM-MinGW toolchain build is used as a benchmark throughout. Instrumentation via bash's PS4 variable revealed that serial configuration phases (cmake, git, configure) were the primary bottleneck limiting CPU utilization to ~53%. Results on AmpereOne 192-core hardware cut the parallel build phase roughly in half compared to Altra 128-core, completing in 603 seconds. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How much does setting the CPU governor to performance mode speed up a compilation-heavy build server? Switching all CPU cores to the performance governor produced a 17% build time speedup in a tested LLVM-MinGW compilation workload, improving from 1,063 to 909 seconds on an Ampere Altra 128-core server. Dynamic governors like ondemand, conservative, or schedutil introduce latency because cores must ramp up frequency after work arrives, which adds up across hundreds of parallel compiler jobs. _Anyone chasing faster CI build times can compare tuning techniques like this through daily.dev._ ### Does lowering CONFIG_HZ improve compilation build times on Linux? Yes, reducing CONFIG_HZ from the default 250 Hz to 100 Hz improved a compilation build from 909.31 to 889.2 seconds, roughly a 2% additional gain. A lower scheduler tick frequency means less frequent preemption checks, which matters because compiler processes are often short-lived (under a second, sometimes under 100ms) and get disproportionately hurt by context-switch overhead. _Engineers tuning kernel parameters for build throughput can track findings like this via daily.dev._ ### Does using a RAM disk for build artifacts meaningfully speed up compiling LLVM-MinGW? Only marginally: building on a 50GB tmpfs RAM disk versus an SSD reduced build time from 1069.1 to 1062.7 seconds, about a 0.6% speedup. The gain is limited because the LLVM-MinGW build's real bottleneck was single-threaded serial phases like configure and cmake, not disk I/O contention. _Developers evaluating whether RAM disks are worth the setup effort for CI builds can find comparisons like this on daily.dev._ ## Similar posts on daily.dev - [ARM Linux Server Performance Up More Than 7x Geo Mean In 8 Years, As Much As 15x With NVIDIA Vera CPU](https://daily.dev/posts/arm-linux-server-performance-up-more-than-7x-geo-mean-in-8-years-as-much-as-15x-with-nvidia-vera-cp-jlxrmmw3i) · Phoronix · 0 upvotes · 0 comments - [Building a Host-Tuned GCC to Make GCC Compile Faster](https://daily.dev/posts/building-a-host-tuned-gcc-to-make-gcc-compile-faster-wt5gmdqdk) · Lobsters · 0 upvotes · 0 comments - [Linux Kernel Build Times Ready To Be Significantly Reduced With Latest Patches](https://daily.dev/posts/linux-kernel-build-times-ready-to-be-significantly-reduced-with-latest-patches-lvdiczew9) · Phoronix · 0 upvotes · 0 comments - [Why We Moved Our Timefold Java Worker Pods from AMD to ARM64](https://daily.dev/posts/why-we-moved-our-timefold-java-worker-pods-from-amd-to-arm64-3vp26e6n0) · Foojay.io · 1 upvotes · 0 comments - [The Async Multiprocessing Has Sound Foundations in Linux, But Further Testing Is Needed](https://daily.dev/posts/the-async-multiprocessing-has-sound-foundations-in-linux-but-further-testing-is-needed-lpf8czcgo) · InfoQ · 0 upvotes · 0 comments --- Tags: [#linux](https://daily.dev/tags/linux), [#cicd](https://daily.dev/tags/cicd), [#arm](https://daily.dev/tags/arm), [#build-systems](https://daily.dev/tags/build-systems) [View this post on daily.dev](https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Optimizing Arm-based Build Servers","url":"https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm"},"datePublished":"2026-07-17T18:29:31.196Z","dateModified":"2026-09-13T20:41:36.197Z","description":"A practical guide to optimizing Linux build servers on Ampere Arm-based hardware (Altra and AmpereOne). Covers enabling parallel builds across all cores, using...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0fcdeb1fc1daa70c22a7db04adea3b5d?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0fcdeb1fc1daa70c22a7db04adea3b5d?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"SitePoint","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"SitePoint","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/sitepoint","url":"https://daily.dev/sources/sitepoint"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"linux,cicd,arm,build-systems","timeRequired":"PT14M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"SitePoint","item":"https://daily.dev/sources/sitepoint"},{"@type":"ListItem","position":3,"name":"Optimizing Arm-based Build Servers"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/optimizing-arm-based-build-servers-rh07w05dm#faq","mainEntity":[{"@type":"Question","name":"How much does setting the CPU governor to performance mode speed up a compilation-heavy build server?","acceptedAnswer":{"@type":"Answer","text":"Switching all CPU cores to the performance governor produced a 17% build time speedup in a tested LLVM-MinGW compilation workload, improving from 1,063 to 909 seconds on an Ampere Altra 128-core server. Dynamic governors like ondemand, conservative, or schedutil introduce latency because cores must ramp up frequency after work arrives, which adds up across hundreds of parallel compiler jobs. Anyone chasing faster CI build times can compare tuning techniques like this through daily.dev."}},{"@type":"Question","name":"Does lowering CONFIG_HZ improve compilation build times on Linux?","acceptedAnswer":{"@type":"Answer","text":"Yes, reducing CONFIG_HZ from the default 250 Hz to 100 Hz improved a compilation build from 909.31 to 889.2 seconds, roughly a 2% additional gain. A lower scheduler tick frequency means less frequent preemption checks, which matters because compiler processes are often short-lived (under a second, sometimes under 100ms) and get disproportionately hurt by context-switch overhead. Engineers tuning kernel parameters for build throughput can track findings like this via daily.dev."}},{"@type":"Question","name":"Does using a RAM disk for build artifacts meaningfully speed up compiling LLVM-MinGW?","acceptedAnswer":{"@type":"Answer","text":"Only marginally: building on a 50GB tmpfs RAM disk versus an SSD reduced build time from 1069.1 to 1062.7 seconds, about a 0.6% speedup. The gain is limited because the LLVM-MinGW build's real bottleneck was single-threaded serial phases like configure and cmake, not disk I/O contention. Developers evaluating whether RAM disks are worth the setup effort for CI builds can find comparisons like this on daily.dev."}}]} ```