{"id":6172,"date":"2026-10-19T16:49:46","date_gmt":"2026-10-19T16:49:46","guid":{"rendered":"https:\/\/i10xnews.kinsta.cloud\/?p=6172"},"modified":"2026-10-19T17:24:54","modified_gmt":"2026-10-19T17:24:54","slug":"turbovla-realtime-edge-vision-language-action","status":"publish","type":"post","link":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action","title":{"rendered":"TurboVLA Runs Robot Policies at 32 Hz on an RTX 4090 Using Under 1 GB VRAM"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">\u26a1 Quick Take<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>The finding:<\/strong> TurboVLA (0.2B params) achieves 97.6% average success on LIBERO with 31.2 ms inference and 0.9 GiB VRAM on a consumer RTX 4090 \u2014 about 32 Hz.<\/li><li><strong>Vs heavy VLAs:<\/strong> Higher average success than \u03c00.5 (96.9%) with ~6% of the parameters and 8\u00d7 faster inference.<\/li><li><strong>Transfer:<\/strong> A 0.4B variant reaches 88.06% on RoboTwin 2.0, matching or beating larger LLM-centric VLAs.<\/li><\/ul>\n\n\n<p class=\"wp-block-paragraph\">LLM-centric vision-language-action policies often sit on remote servers at a few hertz. Huazhong University of Science and Technology researchers cut the LLM out of the inner loop so edge devices can update actions every ~31 ms.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1344\" height=\"768\" src=\"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-edge.webp\" alt=\"Stat card: 97.6 percent success, 31.2 milliseconds latency, 0.9 GiB VRAM, 0.2 billion parameters.\" class=\"wp-image-6171\" srcset=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-edge.webp 1344w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-edge-300x171.webp 300w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-edge-1024x585.webp 1024w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-edge-768x439.webp 768w\" sizes=\"auto, (max-width: 1344px) 100vw, 1344px\" \/><figcaption class=\"wp-element-caption\">i10X chart based on data from Xie et al. (2026), arXiv:2607.27205, Table 1 \/ abstract (arXiv non-exclusive license).<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2c What the study shows<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Direct vision-language interaction produces actions without huge remote models. Real-world fine-tuning uses 4\u00d765 teleoperated demos for 12.5k steps from a LIBERO-pretrained init.<\/p>\n\n\n<ul class=\"wp-block-list\"><li><strong>Edge economics:<\/strong> Consumer GPU, <1 GB inference memory, no network round-trip.<\/li><li><strong>Deployment contrast:<\/strong> Authors contrast 239 ms remote LLM-centric perception vs 31 ms on-device.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u2705 Why it matters: robots need local loops<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Network-dependent VLAs fail offline and lag physically. Sub-1 GB, 32 Hz policies unlock real edge deployment.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83e\uddea Method in brief<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Lightweight VLA without LLM-centric inner loop; LIBERO and RoboTwin 2.0 evals; real-world fine-tune.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u2753 What remains open<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Benchmark suites:<\/strong> Sim success may not equal open-world robustness.<\/li><li><strong>Task coverage:<\/strong> LIBERO\/RoboTwin scope.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\ud83e\udded Context<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Related embodied AI in <a href=\"\/news\/category\/studies\/multimodal\">Multimodal<\/a>.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udee0\ufe0f What it means for robot startups<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Profile Hz and VRAM:<\/strong> Not only success rate.<\/li><li><strong>Keep LLM out of the control loop:<\/strong> Use it for high-level goals.<\/li><li><strong>Fine-tune on little demo data:<\/strong> 4\u00d765 trajectories reported.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u270d\ufe0f About the analysis<\/h2>\n\n\n<p class=\"wp-block-paragraph\">This article is based on the full text of the preprint &#8220;TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with &lt;1 GB VRAM&#8221; (arXiv:2607.27205v3, submitted July 29, 2026). All numbers are taken from the paper&#8217;s text, tables and figures; i10X did not rerun the experiments. The paper is distributed under arXiv&#8217;s non-exclusive license, so i10X does not reproduce its figures; the chart was drawn by i10X from numbers reported in Table 1 \/ abstract.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2d i10x Perspective<\/h2>\n\n\n<p class=\"wp-block-paragraph\">TurboVLA is a reminder that embodiment is a systems problem. 97.6% means little if the policy updates four times a second on a rented A100. Thirty-two hertz on a 4090 is the more radical number.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>HUST&#8217;s 0.2B TurboVLA hits 97.6% LIBERO success at 31.2 ms latency and 0.9 GiB VRAM \u2014 ~8\u00d7 faster than \u03c00.5 with ~6% of the parameters.<\/p>\n","protected":false},"author":2,"featured_media":6322,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"paper_title":"TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with &lt;1 GB VRAM","paper_url":"https:\/\/arxiv.org\/abs\/2607.27205","pdf_url":"https:\/\/arxiv.org\/pdf\/2607.27205v3","doi":"","arxiv_id":"2607.27205","authors":"Hengyi Xie, Chenfei Yao, Xianjin Wu, Yingying Zhu, Dingkang Liang, Xiang Bai, Han Ding","institutions":"Huazhong University of Science and Technology","paper_date":"2026-07-29","venue":"arXiv (cs.CV), version 3","peer_reviewed":false,"code_url":"https:\/\/github.com\/H-EmbodVis\/TurboVLA","data_url":"","funding_coi":"Academic lab.","figure_license":"Chart drawn by i10X from Table 1 \/ abstract (arXiv non-exclusive license)","paper_license":"arXiv non-exclusive license","replication":"no","study_models":"TurboVLA 0.2B\/0.4B vs \u03c00.5 and larger LLM-centric VLAs","study_sample":"LIBERO, RoboTwin 2.0, real-world teleop fine-tune","footnotes":"","_yoast_wpseo_primary_category":""},"categories":[58,52],"tags":[],"companies":[],"products":[],"class_list":["post-6172","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-multimodal","category-studies"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.7 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News<\/title>\n<meta name=\"description\" content=\"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News\" \/>\n<meta property=\"og:description\" content=\"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action\" \/>\n<meta property=\"og:site_name\" content=\"i10X News\" \/>\n<meta property=\"article:published_time\" content=\"2026-10-19T16:49:46+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-10-19T17:24:54+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1344\" \/>\n\t<meta property=\"og:image:height\" content=\"768\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Christopher Ort\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:site\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Christopher Ort\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"1 minute\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action\"},\"author\":{\"name\":\"Christopher Ort\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\"},\"headline\":\"TurboVLA Runs Robot Policies at 32 Hz on an RTX 4090 Using Under 1 GB VRAM\",\"datePublished\":\"2026-10-19T16:49:46+00:00\",\"dateModified\":\"2026-10-19T17:24:54+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action\"},\"wordCount\":144,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/turbovla-realtime-edge-vision-language-action-title.webp\",\"articleSection\":[\"Multimodal\",\"Studies\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action\",\"name\":\"TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/turbovla-realtime-edge-vision-language-action-title.webp\",\"datePublished\":\"2026-10-19T16:49:46+00:00\",\"dateModified\":\"2026-10-19T17:24:54+00:00\",\"description\":\"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#primaryimage\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/turbovla-realtime-edge-vision-language-action-title.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/turbovla-realtime-edge-vision-language-action-title.webp\",\"width\":1344,\"height\":768,\"caption\":\"Illustration: i10X\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/turbovla-realtime-edge-vision-language-action#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/news.i10x.ai\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"TurboVLA Runs Robot Policies at 32 Hz on an RTX 4090 Using Under 1 GB VRAM\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"name\":\"i10X News\",\"description\":\"Stay updated with the latest news and announcements from i10X.\",\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/news.i10x.ai\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\",\"name\":\"i10X\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"contentUrl\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"width\":180,\"height\":180,\"caption\":\"i10X\"},\"image\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/i10x_ai\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/i10x-ai\",\"https:\\\/\\\/instagram.com\\\/i10x.ai\",\"https:\\\/\\\/www.youtube.com\\\/@i10X_ai_official\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\",\"name\":\"Christopher Ort\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"caption\":\"Christopher Ort\"},\"description\":\"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/author\\\/christopher-ort\",\"sameAs\":[\"https:\\\/\\\/i10x.ai\\\/blog\\\/author\\\/christopher-ort\"],\"worksFor\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"alumniOf\":{\"@type\":\"EducationalOrganization\",\"name\":\"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)\",\"url\":\"https:\\\/\\\/www.fau.eu\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News","description":"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action","og_locale":"en_US","og_type":"article","og_title":"TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News","og_description":"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.","og_url":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action","og_site_name":"i10X News","article_published_time":"2026-10-19T16:49:46+00:00","article_modified_time":"2026-10-19T17:24:54+00:00","og_image":[{"width":1344,"height":768,"url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp","type":"image\/webp"}],"author":"Christopher Ort","twitter_card":"summary_large_image","twitter_creator":"@i10x_ai","twitter_site":"@i10x_ai","twitter_misc":{"Written by":"Christopher Ort","Est. reading time":"1 minute"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#article","isPartOf":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action"},"author":{"name":"Christopher Ort","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a"},"headline":"TurboVLA Runs Robot Policies at 32 Hz on an RTX 4090 Using Under 1 GB VRAM","datePublished":"2026-10-19T16:49:46+00:00","dateModified":"2026-10-19T17:24:54+00:00","mainEntityOfPage":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action"},"wordCount":144,"commentCount":0,"publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"image":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp","articleSection":["Multimodal","Studies"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#respond"]}]},{"@type":"WebPage","@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action","url":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action","name":"TurboVLA: 32 Hz Vision-Language-Action on a 4090 - i10X News","isPartOf":{"@id":"https:\/\/news.i10x.ai\/#website"},"primaryImageOfPage":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#primaryimage"},"image":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp","datePublished":"2026-10-19T16:49:46+00:00","dateModified":"2026-10-19T17:24:54+00:00","description":"TurboVLA: 97.6% LIBERO, 31 ms latency, &lt;1 GB VRAM on RTX 4090. 0.4B variant reaches 88% on RoboTwin 2.0.","breadcrumb":{"@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#primaryimage","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/turbovla-realtime-edge-vision-language-action-title.webp","width":1344,"height":768,"caption":"Illustration: i10X"},{"@type":"BreadcrumbList","@id":"https:\/\/i10x.ai\/news\/turbovla-realtime-edge-vision-language-action#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/news.i10x.ai\/"},{"@type":"ListItem","position":2,"name":"TurboVLA Runs Robot Policies at 32 Hz on an RTX 4090 Using Under 1 GB VRAM"}]},{"@type":"WebSite","@id":"https:\/\/news.i10x.ai\/#website","url":"https:\/\/news.i10x.ai\/","name":"i10X News","description":"Stay updated with the latest news and announcements from i10X.","publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/news.i10x.ai\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/news.i10x.ai\/#organization","name":"i10X","url":"https:\/\/news.i10x.ai\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/","url":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","contentUrl":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","width":180,"height":180,"caption":"i10X"},"image":{"@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/i10x_ai","https:\/\/www.linkedin.com\/company\/i10x-ai","https:\/\/instagram.com\/i10x.ai","https:\/\/www.youtube.com\/@i10X_ai_official"]},{"@type":"Person","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a","name":"Christopher Ort","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","caption":"Christopher Ort"},"description":"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.","url":"https:\/\/i10x.ai\/news\/author\/christopher-ort","sameAs":["https:\/\/i10x.ai\/blog\/author\/christopher-ort"],"worksFor":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"alumniOf":{"@type":"EducationalOrganization","name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)","url":"https:\/\/www.fau.eu"}}]}},"_links":{"self":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6172","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/comments?post=6172"}],"version-history":[{"count":2,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6172\/revisions"}],"predecessor-version":[{"id":6323,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6172\/revisions\/6323"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media\/6322"}],"wp:attachment":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media?parent=6172"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/categories?post=6172"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/tags?post=6172"},{"taxonomy":"company","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/companies?post=6172"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/products?post=6172"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}