--- title: Lemonade 11.9 Local AI Server Released With Super... description: Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds an experimental llamacpp-hrx backend built on AMD's new HRX... canonical: https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Lemonade 11.9 Local AI Server Released With Super Exciting AMD ROCm HRX Backend | daily.dev og:description: Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds an experimental llamacpp-hrx backend built on AMD's new HRX... og:url: https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt og:image: https://api.daily.dev/og/posts/OQelo2sDT.png og:image:alt: Lemonade 11.9 Local AI Server Released With Super Exciting AMD ROCm HRX Backend 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. # Lemonade 11.9 Local AI Server Released With Super Exciting AMD ROCm HRX Backend **[Phoronix](https://daily.dev/sources/phoronix)** · 4 min read · 0 upvotes · 0 comments ## Summary Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds an experimental llamacpp-hrx backend built on AMD's new HRX runtime, targeting Radeon RX 7900 series (RDNA3) and Strix Halo APUs on Linux (GFX1100 and GFX1151). HRX is a lighter, client-focused subset of ROCm developed with AMD's Loom/Hyperloom compiler stack, offering a native alternative to Vulkan and HIP backends. AMD engineer Stella Laurenzo detailed the effort in a llama.cpp GitHub discussion, citing 30-50% prefill token/s uplift and up to 10% decode uplift versus existing backends, though the work remains early and limited to a minimal kernel set for the Qwen3-30B-A3B model. Lemonade 11.9 also adds Qwen3-Next support to its pinned llama.cpp backend. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is AMD's HRX backend for llama.cpp and which GPUs does it support? HRX is a new, lighter subset of ROCm co-developed by AMD engineers specifically to give llama.cpp a native backend independent of the existing Vulkan and HIP backends. The initial experimental support targets Radeon RX 7900 series (RDNA3, GFX1100) and Strix Halo APUs (RDNA3.5, GFX1151) on Linux, and ships in Lemonade version 11.9.0. _Developers optimizing local LLM inference on AMD hardware can follow backend changes like this via daily.dev._ ### How much faster is AMD's HRX backend compared to existing llama.cpp Vulkan or HIP backends? AMD engineers reported a possible 30 to 50 percent tokens-per-second uplift on prefill compared to llama.cpp's existing Vulkan and HIP backends, with up to roughly 10 percent uplift on non-MTP decode tasks. The backend is still experimental, tested initially on the unsloth Qwen3-30B-A3B-Instruct-2507-GGUF Q4_K_M model on RDNA3 (W7900) and RDNA3.5 (Strix Halo) hardware. _Teams evaluating AMD versus other inference backends can track these benchmarks on daily.dev._ ### What new features does Lemonade 11.9 add besides the HRX backend? Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds Qwen3-Next support to its pinned llama.cpp backend in addition to the experimental llamacpp-hrx backend for AMD GFX1100 and GFX1151 GPU targets on Linux. _Anyone running local AI servers can keep pace with releases like this through daily.dev._ ## Similar posts on daily.dev - [AMD's Lemonade SDK For Local AI Adds NVIDIA CUDA Support](https://daily.dev/posts/amd-s-lemonade-sdk-for-local-ai-adds-nvidia-cuda-support-gkd6dj05w) · Phoronix · 1 upvotes · 0 comments - [I switched my local AI setup to AMD's Lemonade after Nvidia support landed, and solved my local AI portability problem](https://daily.dev/posts/i-switched-my-local-ai-setup-to-amd-s-lemonade-after-nvidia-support-landed-and-solved-my-local-ai-p-ede18yoda) · XDA Developers · 1 upvotes · 0 comments - [AMD Releases Lemonade 11.0 Local AI Server With Text-To-Speech, Other New Features](https://daily.dev/posts/amd-releases-lemonade-11-0-local-ai-server-with-text-to-speech-other-new-features-mkmvvd9dt) · Phoronix · 2 upvotes · 0 comments --- Tags: [#gpu](https://daily.dev/tags/gpu), [#amd](https://daily.dev/tags/amd), [#llama-cpp](https://daily.dev/tags/llama-cpp), [#rocm](https://daily.dev/tags/rocm) [View this post on daily.dev](https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt) ```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":"Lemonade 11.9 Local AI Server Released With Super Exciting AMD ROCm HRX Backend","url":"https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt"},"datePublished":"2026-09-03T00:36:34.156Z","dateModified":"2026-09-03T02:46:11.320Z","description":"Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds an experimental llamacpp-hrx backend built on AMD's new HRX...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/ff81b7af3b7b91a6d67e7da41f4860a6?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/ff81b7af3b7b91a6d67e7da41f4860a6?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Phoronix","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":"Phoronix","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/6a26ba379dd64932830aa0c9803dd7ee","url":"https://daily.dev/sources/phoronix"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"gpu,amd,llama-cpp,rocm","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Phoronix","item":"https://daily.dev/sources/phoronix"},{"@type":"ListItem","position":3,"name":"Lemonade 11.9 Local AI Server Released With Super Exciting AMD ROCm HRX Backend"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/lemonade-11-9-local-ai-server-released-with-super-exciting-amd-rocm-hrx-backend-oqelo2sdt#faq","mainEntity":[{"@type":"Question","name":"What is AMD's HRX backend for llama.cpp and which GPUs does it support?","acceptedAnswer":{"@type":"Answer","text":"HRX is a new, lighter subset of ROCm co-developed by AMD engineers specifically to give llama.cpp a native backend independent of the existing Vulkan and HIP backends. The initial experimental support targets Radeon RX 7900 series (RDNA3, GFX1100) and Strix Halo APUs (RDNA3.5, GFX1151) on Linux, and ships in Lemonade version 11.9.0. Developers optimizing local LLM inference on AMD hardware can follow backend changes like this via daily.dev."}},{"@type":"Question","name":"How much faster is AMD's HRX backend compared to existing llama.cpp Vulkan or HIP backends?","acceptedAnswer":{"@type":"Answer","text":"AMD engineers reported a possible 30 to 50 percent tokens-per-second uplift on prefill compared to llama.cpp's existing Vulkan and HIP backends, with up to roughly 10 percent uplift on non-MTP decode tasks. The backend is still experimental, tested initially on the unsloth Qwen3-30B-A3B-Instruct-2507-GGUF Q4_K_M model on RDNA3 (W7900) and RDNA3.5 (Strix Halo) hardware. Teams evaluating AMD versus other inference backends can track these benchmarks on daily.dev."}},{"@type":"Question","name":"What new features does Lemonade 11.9 add besides the HRX backend?","acceptedAnswer":{"@type":"Answer","text":"Lemonade 11.9, an AMD-backed open-source local AI server for Linux, Windows, and macOS, adds Qwen3-Next support to its pinned llama.cpp backend in addition to the experimental llamacpp-hrx backend for AMD GFX1100 and GFX1151 GPU targets on Linux. Anyone running local AI servers can keep pace with releases like this through daily.dev."}}]} ```