{"id":82,"date":"2026-11-16T04:16:00","date_gmt":"2026-11-16T04:16:00","guid":{"rendered":"https:\/\/blockoala.com\/index.php\/2026\/11\/16\/harmonyos-7-fast-kit-tested-faster-sorting-with-native-integration-costs\/"},"modified":"2026-11-16T04:16:00","modified_gmt":"2026-11-16T04:16:00","slug":"harmonyos-7-fast-kit-tested-faster-sorting-with-native-integration-costs","status":"publish","type":"post","link":"https:\/\/blockoala.com\/index.php\/2026\/11\/16\/harmonyos-7-fast-kit-tested-faster-sorting-with-native-integration-costs\/","title":{"rendered":"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs"},"content":{"rendered":"<p><em>Originally reported November 12, 2026.<\/em><\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/Fmgco55uSHZvmQpSkGzYM.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (1)\"><\/figure>\n<p>An <a href=\"https:\/\/www.infoq.cn\/article\/kIw9Wp2uSgCwrsgicIZ0\" rel=\"noopener nofollow\" target=\"_blank\">InfoQ hands-on test<\/a> examined two HarmonyOS 7 FAST Kit capabilities: natural-language sorting and a 4,096-point FFT for spectrum analysis. Both use Native C interfaces, so an ArkTS application needs a NAPI bridge. The tests ran in a Mate 90 Pro simulator on HarmonyOS 7.0.0.107, using DevEco Studio 26.0.0.851 and the HarmonyOS 26.0.0 Release SDK\u2014not on a physical phone.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/O7D0widocmTmVnvvAa4nB.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (2)\"><\/figure>\n<h3>Sorting: a faster result in this test<\/h3>\n<p>The article compared FAST Kit\u2019s <code>HMS_FAST_Algo_NaturalSort<\/code> with an ArkTS <code>Intl.Collator<\/code> baseline across lists of 1,000, 10,000 and 100,000 entries. In one run, sorting 10,000 entries took 20 ms with FAST Kit versus 146 ms with the baseline. Across the reported tests, the measured speedup was 7.3\u201313.3 times; at 1,000 entries, the absolute difference was 37 ms. The article says the sorting-result hashes matched in every run.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/qAFUe8wwZzFNzYDwhSFm3.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (3)\"><\/figure>\n<p>A separate deployment through the <code>hdc<\/code> command line recorded 51\/262\/2,381 ms for the baseline versus 4\/30\/203 ms for FAST Kit at the three respective list sizes. These results should not be equated with the 60%\u201370% improvement over a third-party internationalization library cited elsewhere in the article: the baselines and execution paths differ.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/4eLy4l88OVUqbUNBLlFyG.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (4)\"><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/vXcYUkKNb7C67DD9sYqd_-scaled.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (5)\"><\/figure>\n<h3>FFT: fast computation is only part of the work<\/h3>\n<p>For the second experiment, FAST Kit processed simulated signals with a 4,096-point real FFT. The article reports approximately 0.02\u20130.03 ms per transform in the simulator. For a generated three-tone signal with amplitudes of 0.60, 0.30 and 0.15, the detected amplitudes were 0.600, 0.300 and 0.150.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/xyUQVOZ4MckO9slazViZw-scaled.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (6)\"><\/figure>\n<p>The initial spectrum chart nevertheless showed no bars, despite detecting peaks at 50, 120 and 300 Hz. According to the article, those frequencies fell exactly on FFT bins, while the chart\u2019s sampled bar positions missed the concentrated peaks. Adding a Hann window made them visible, but required amplitude compensation. Windowing, amplitude calibration and chart sampling remain application responsibilities; transform time alone does not establish end-to-end performance.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/nYMXstqY5mvYuy8DSzTIJ-scaled.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (7)\"><\/figure>\n<h3>What developers need to account for<\/h3>\n<p>The tested interfaces require API 26, Native-module integration and compatible SDK configuration. The article specifies <code>compatibleSdkVersion<\/code> as <code>\"26.0.0\"<\/code> and notes that linking the FAST Kit libraries requires the HMS sysroot. It also recommends validating inputs, testing on physical devices at the minimum supported version, and providing a fallback where the system libraries are unavailable. Those steps matter particularly for smaller workloads, where integration cost may outweigh the measured time saving.<\/p>\n<figure><img decoding=\"async\" src=\"https:\/\/blockoala.com\/wp-content\/uploads\/2026\/11\/uCO20nVtoEPTh9ayKsPMs-scaled.png\" alt=\"HarmonyOS 7 FAST Kit tested: faster sorting, with Native integration costs (8)\"><\/figure>\n<p>Source: <a href=\"https:\/\/www.infoq.cn\/article\/kIw9Wp2uSgCwrsgicIZ0\" rel=\"noopener nofollow\" target=\"_blank\">infoq.cn<\/a> (originally published 2026-11-12 10:20 UTC+8)<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Originally reported November 12, 2026. An InfoQ hands-on test examined two HarmonyOS 7 FAST Kit capabilities: natural-language sorting [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":64,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-82","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/posts\/82","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/comments?post=82"}],"version-history":[{"count":0,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/posts\/82\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/media\/64"}],"wp:attachment":[{"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/media?parent=82"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/categories?post=82"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blockoala.com\/index.php\/wp-json\/wp\/v2\/tags?post=82"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}