#strixhalo — Public Fediverse posts

Yeah this is the model I have been wanting for my #StrixHalo since I got it. Perfect.

Два AMD Strix Halo в AI‑инфраструктуре: 34 контейнера на одном, ~70 tok/s Qwen3.6 на другом

На узле моей AI‑платформы крутятся 34 контейнера: Dify, RAGFlow, векторные базы, мониторинг и SSO. Большой языковой модели среди них нет: основную генерацию стек получает по LAN с DGX Spark. На втором таком же мини‑ПК я отдельно поднял локальную Qwen3.6–35B‑A3B и прогнал серию замеров от 1K до 64K при контекстном окне 256K. Обе машины — Beelink GTR9 Pro на Ryzen AI Max+ 395 (Strix Halo). Ниже — что эти коробки реально умеют: 117,4 ГиБ GTT после настройки ttm.pages_limit , p50/p95 локальных эмбеддингов и реранка, около 70 tok/s генерации через Vulkan/RADV и три грабли gfx1151.

https://habr.com/ru/articles/1058502/

#strixhalo #amd #selfhosted #llm #vulkan #rocm

Come Scegliere il Miglior Mini PC per l'AI Locale nel 2026: Strix Halo vs DGX Spark vs Mac

Un mini PC grande quanto un libro tascabile è oggi in grado di eseguire localmente un modello da 200 miliardi di parametri. Ma la scelta non dipende solo dal prezzo. La capacità della memoria determina quali modelli possono essere caricati, la larghezza di banda influisce sulla velocità di esecuzione e lo stack software — CUDA, ROCm o Metal — stabilisce se gli strumenti che utilizi funzioneranno davvero. Ecco un confronto tra le quattro principali opzioni disponibili nel 2026, con prezzi e benchmark aggiornati.

buysellram.com/blog/how-to-cho…

#AIlocale #LLM #MiniPC #HardwareAI #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #InfrastrutturaAI #Ollama #RyzenAI #AIPC #AMD #NVIDIA #Apple

Come Scegliere il Miglior Mini PC per l'AI Locale nel 2026: Strix Halo vs DGX Spark vs Mac

Un mini PC grande quanto un libro tascabile è oggi in grado di eseguire localmente un modello da 200 miliardi di parametri. Ma la scelta non dipende solo dal prezzo. La capacità della memoria determina quali modelli possono essere caricati, la larghezza di banda influisce sulla velocità di esecuzione e lo stack software — CUDA, ROCm o Metal — stabilisce se gli strumenti che utilizi funzioneranno davvero. Ecco un confronto tra le quattro principali opzioni disponibili nel 2026, con prezzi e benchmark aggiornati.

buysellram.com/blog/how-to-cho…

#AIlocale #LLM #MiniPC #HardwareAI #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #InfrastrutturaAI #Ollama #RyzenAI #AIPC #AMD #NVIDIA #Apple

Lilbits: Flipper Zero’s firmware future, a new Linux gaming laptop, an E Ink monitor, and AMD’s pricey Ryzen Halo mini PC

Shortly after launching a new thin and light Linux laptop with an Intel Panther Lake processor, Linux PC company System76 is preparing to launch… another notebook that could be described the same way.

But the updated System76 Adder Pro also packs features like discrete graphics and an OLED display with a high refresh rate, which means that it’ll both be a better fit for gaming or mobile […]

The AMD Ryzen Halo mini workstation with Ryzen AI MAX+ 395, 128GB LPDDR5x RAM, a 2TB SSD, 10 Gigabit Ethernet, and Windows and Linux support is now available from Micro Center... for $4000. https://www.amd.com/en/blogs/2026/amd-ryzen-ai-halo-now-available-at-micro-center.html #RyzenHalo #MiniPC #StrixHalo #AMD

The AMD Ryzen Halo mini workstation with Ryzen AI MAX+ 395, 128GB LPDDR5x RAM, a 2TB SSD, 10 Gigabit Ethernet, and Windows and Linux support is now available from Micro Center... for $4000. https://www.amd.com/en/blogs/2026/amd-ryzen-ai-halo-now-available-at-micro-center.html #RyzenHalo #MiniPC #StrixHalo #AMD

Great DGX upsell, btw.

https://videocardz.com/newz/amd-launches-3999-ryzen-ai-halo-pc-limited-to-one-us-retailer

#AMD #Ryzen #RyzenAI #StrixHalo #Zen5 #RDNA3 #AI #LLM #ArtificialIntelligence

Great DGX upsell, btw.

https://videocardz.com/newz/amd-launches-3999-ryzen-ai-halo-pc-limited-to-one-us-retailer

#AMD #Ryzen #RyzenAI #StrixHalo #Zen5 #RDNA3 #AI #LLM #ArtificialIntelligence

Cringe model card image but actually pretty impressive Stepfun 3.7 Flash quant for #StrixHalo

https://huggingface.co/jcbtc/Step-3.7-Flash-ROCmFPX-Q3-QualityPlus

GMK EVO-X3 with Ryzen AI Max+ 395 and 128GB RAM now available for $3600 and up

The GMK EVO-X3 is a slim desktop computer that packs a lot of power into a compact design. With an AMD Ryzen AI Max+ 395 processor featuring Radeon 8060S discrete-class graphics, 128GB of RAM with 256GB/s bandwidth, and a robust set of ports including USB4 and OCuLink connectors, GMK is positioning the EVO-X3 as an AI […]

GPD Win Max 3 is a mini laptop for work and play with AMD Strix Halo and a removable battery. with a removable

The GPD Win Max 3 is a mini-laptop that packs a lot of horsepower into a compact design. With a 9.06 inch display and a nearly full-sized keyboard and Precision touchpad, it’s a little laptop you can use for productivity on the go. But remove a couple of magnetic panels above the keyboard and you’ll find a set of joysticks for gaming.

And with an AMD Strix Halo processor featuring […]

Cool guide for getting #RDMA working for AMD #StrixHalo in #Linux. Reminds me of a startup I was in ~15 years ago where we used #Infiniband for #GlusterFS. Looks like it still needs some work to remain stable. The troubleshooting section has a deadlock warning (and fix) right off the bat.
https://github.com/kyuz0/amd-strix-halo-vllm-toolboxes/blob/main/rdma_cluster/setup_guide.md

Cool guide for getting #RDMA working for AMD #StrixHalo in #Linux. Reminds me of a startup I was in ~15 years ago where we used #Infiniband for #GlusterFS. Looks like it still needs some work to remain stable. The troubleshooting section has a deadlock warning (and fix) right off the bat.
https://github.com/kyuz0/amd-strix-halo-vllm-toolboxes/blob/main/rdma_cluster/setup_guide.md

Oh boy another ROCM llama.cpp fork for #StrixHalo ​:ExhaustedPepe:​
https://github.com/charlie12345/ROCmFPX

GMK EVO-X3 mini PC with Ryzen AI Max+ 395 and up to 128GB RAM launches this month

The GMK EVO-X3 is a compact workstation with a 16-core, 32-thread processor, discrete-class integrated graphics, and plenty of I/O including an OCuLink port for an external PCIe 4.0 connection to a graphics dock or other add-ons.

First revealed earlier this year, the EVO-X3 will be available for “early access registration” on June 22, ahead of a global launch on June 29th July 6th. (Update: […]

I have my own issue with #Deepseek v4 Pro in terms of performance to params. Though I do recognize how in terms of the tech behind it how much of a achievement it is.

So when I first heard about the ds4 project I scoffed and thought wow that is going to run like ass ass. Though after hearing about it more and more in #StrixHalo circles I figured why not give it a shot and I must say I am quite impressed with what it does on low power hardware. It is not major ground breaking but it is worth having as a option if needed. Was even able to link of the jank ds4-server to my llama-swap config.

Even appreciate the work Donato Capitella added to the Strix Halo cause as always/ (Even if I have no need for his toolboxes)
https://youtu.be/Cfl3TS7ME5s

Recommended: A new #deepseek v4 toolbox over at https://strix-halo-toolboxes.com/
lets me run a powerful DeepSeek v4 flash Q4 quant LLM on dual Strix Halo 128GB. Give it a try! #strixhalo #localAI #inference

A mini PC the size of a paperback can now run a 200B-parameter model locally. But choosing one isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #RyzenAI #AIPC #AMD #NVIDIA #Apple

A mini PC the size of a paperback can now run a 200B-parameter model locally. But choosing one isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #RyzenAI #AIPC #AMD #NVIDIA #Apple

Choosing A mini PC isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #AIinfrastructure #Ollama #RyzenAI #AIPC #AMD #NVIDIA

Choosing A mini PC isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #AIinfrastructure #Ollama #RyzenAI #AIPC #AMD #NVIDIA

AMD Ryzen AI Max+ 395 vs Nvidia DGX Spark vs Apple Mac — plus when a GPU tower still beats all three. A practical hardware guide for IT managers, developers, and small-business owners weighing a local LLM machine.

Running large language models locally went from a niche hobby to a real procurement question in 2026. A mini PC the size of a paperback can now hold a 200-billion-parameter model — the kind of workload that used to need a server rack.

But picking one isn't about the lowest price. Three things decide whether a model runs well: memory capacity (what fits), memory bandwidth (how fast it runs), and the software ecosystem — CUDA, ROCm, or Metal — that determines whether your existing tools work at all.

There are four real ways to run a local LLM on your desk: a discrete-GPU tower (fastest, but a VRAM wall), AMD Strix Halo mini PCs (big unified memory, cheap, Windows-native), Nvidia's GB10 boxes like the DGX Spark and Dell Pro Max (CUDA, but now $4,699 and Linux-only), and Apple's Mac mini and Mac Studio (high bandwidth, silent, no CUDA).

This guide breaks down which fits which job — with verified specs and current prices.

An appendix at the end collects what early buyers of the AMD “lunchbox” are actually reporting.

https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #AIinfrastructure #Ollama #RyzenAI #AIPC #AMD #NVIDIA #Apple #technology

AMD Ryzen AI Max+ 395 vs Nvidia DGX Spark vs Apple Mac — plus when a GPU tower still beats all three. A practical hardware guide for IT managers, developers, and small-business owners weighing a local LLM machine.

Running large language models locally went from a niche hobby to a real procurement question in 2026. A mini PC the size of a paperback can now hold a 200-billion-parameter model — the kind of workload that used to need a server rack.

But picking one isn't about the lowest price. Three things decide whether a model runs well: memory capacity (what fits), memory bandwidth (how fast it runs), and the software ecosystem — CUDA, ROCm, or Metal — that determines whether your existing tools work at all.

There are four real ways to run a local LLM on your desk: a discrete-GPU tower (fastest, but a VRAM wall), AMD Strix Halo mini PCs (big unified memory, cheap, Windows-native), Nvidia's GB10 boxes like the DGX Spark and Dell Pro Max (CUDA, but now $4,699 and Linux-only), and Apple's Mac mini and Mac Studio (high bandwidth, silent, no CUDA).

This guide breaks down which fits which job — with verified specs and current prices.

An appendix at the end collects what early buyers of the AMD “lunchbox” are actually reporting.

https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #AppleSilicon #EdgeAI #AIinfrastructure #Ollama #RyzenAI #AIPC #AMD #NVIDIA #Apple #technology

ONEXPLAYER X2 Mini Pro handheld gaming PC with Ryzen AI Max+ 388 hits Indiegogo for $2466 and up

The ONEXPLAYER X2 Mini Pro is a handheld game console with an 8.8 inch, 1920 x 1200 pixel, 30 -144 Hz AMOLED display, detachable controllers, and the most powerful graphics currently available for a device this size.

But with an AMD Strix Halo processor and at least 48GB of high-speed memory inside, it’s unsurprising that this handheld is expensive. The ONEXPLAYER X2 Mini Pro is now available […]

A mini PC the size of a paperback can now run a 200B-parameter model locally. But choosing one isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #EdgeAI #AIinfrastructure #tech #RyzenAI #NVIDIA #Apple

A mini PC the size of a paperback can now run a 200B-parameter model locally. But choosing one isn't about the lowest price tag. Capacity sets what fits, bandwidth sets how fast it runs, and the software stack — CUDA, ROCm, or Metal — decides whether your tools work at all. Here's how the four real options compare in 2026, with current prices and benchmarks.
https://www.buysellram.com/blog/how-to-choose-the-best-mini-pc-for-local-ai-in-2026/
#LocalAI #LLM #MiniPC #AIhardware #StrixHalo #DGXSpark #EdgeAI #AIinfrastructure #tech #RyzenAI #NVIDIA #Apple

Neato that video gen worked. 35min~ is not bad for the #StrixHalo box. Not sure how much i will use it.....had more fun learning all the inner working s and getting it to work than I did actually gening the video

#LTX

MINISFORUM N5 MAX NAS with AMD Strix Halo launches for $2469

The MINISFORUM N5 Max is a computer with an AMD Ryzen AI Max+ 395 Strix Halo processor featuring a 16-core, 32-thread CPU, a discrete-class Radeon 8060S GPU, and high-bandwidth memory.

It’s also a NAS with support for up to 10 storage devices: 5 hard dries and 5 SSDs, as well as dual 10 Gigbit LAN ports, among other features. First unveiled during CES in January, the N5 Max is now available […]

GMK EVO-X3 is a mini workstation with up to Ryzen AI Max+ PRO 495, 192GB of RAM, and OCuLink

The GMK EVO-X3 is an upcoming computer that’s taller and thinner than last year’s EVO-X2, but which brings a few significant upgrades including an OCuLink port and a second M.2 slot that allows you to twice as much storage inside.

GMK also plans to offer at least two processor options. According to wccftech, the GMK EVO-X3 will launch with the same AMD Ryzen AI Max+ 395 chip used in the older […]

I finally loaded a 120B model - #nemotron3 super, onto my #DGXSpark. With all the stars aligned and goats sacrificed, I think this is the NVFP4 flavour. I'm using it to review patches I made earlier for evaluation.

So far I'm blown away by how _fast_ it is, I'm seeing ~20-25 tokens per second.

It's too soon if this is going to replace my go-to model (qwen3.6-35B-A3B) but I'm looking forward to using during my day job tasks.

I run two models, one on #StrixHalo and one more on the spark. A/B :)

I finally loaded a 120B model - #nemotron3 super, onto my #DGXSpark. With all the stars aligned and goats sacrificed, I think this is the NVFP4 flavour. I'm using it to review patches I made earlier for evaluation.

So far I'm blown away by how _fast_ it is, I'm seeing ~20-25 tokens per second.

It's too soon if this is going to replace my go-to model (qwen3.6-35B-A3B) but I'm looking forward to using during my day job tasks.

I run two models, one on #StrixHalo and one more on the spark. A/B :)

Lilbits: Surface Pro with Snapdragon X2 Elite, another Strix Halo mini PC, and more Linux-like tools for Windows

This week Microsoft unveiled the Surface RTX Dev Box mini PC with an NVIDIA RTX Spark processor and a Surface Laptop Ultra with the same chip. Both are expected to arrive later this year and both are aimed squarely at developers. And last month the company introduced business-class Surface Pro and Surface Laptop models powered by Intel Panther Lake processors.

So what’s next? According to […]

The Peladn YO2 is yet another AMD Ryzen AI Max+ 395 mini PC with a high price tag (over $3000), premium specs, and an unusual design (it looks more like some sort of appliance than a PC, with a rounded top and RGB light strip on the front). https://www.notebookcheck.net/New-Peladn-mini-PC-releases-with-AMD-Strix-Halo-and-up-to-96-GB-VRAM-already-being-sold-globally.1314319.0.html #Peladn #PeladnYO2 #MiniPC #StrixHalo

The Peladn YO2 is yet another AMD Ryzen AI Max+ 395 mini PC with a high price tag (over $3000), premium specs, and an unusual design (it looks more like some sort of appliance than a PC, with a rounded top and RGB light strip on the front). https://www.notebookcheck.net/New-Peladn-mini-PC-releases-with-AMD-Strix-Halo-and-up-to-96-GB-VRAM-already-being-sold-globally.1314319.0.html #Peladn #PeladnYO2 #MiniPC #StrixHalo

Acer Veriton RA110 is a mini workstation with Ryzen AI Max+ 395

Acer is joining the Ryzen AI Max+ 395 mini PC bandwagon with the new Veriton RA110. The company is positioning the little computer as a “mini workstation” thanks to its combination of a 16-core, 32-thread processor with 40-core Radeon 8060S discrete-class integrated graphics and support for up to 128GB of LPDDR5x-8533 memory with 256GB/s bandwidth.

But in addition to functioning as a graphics […]

RT @TeksEdge: 🚀 Neue MTP-Unterstützung für Strix Halo veröffentlicht!

mehr auf Arint.info

#AI #AMD #MTP #Qwen #ROCm #StrixHalo #arint_info

https://x.com/TeksEdge/status/2058728175388262761#m

I'm currently using the same model (Qwen3.6-35B-A3B-MTP-GUFF) across my three AI-capable machines: the #StrixHalo Asus PX13, the 2019 #MacPro with Vega 2, and the M2 Max #MacStudio.

The fastest one is ... Vega 2 with 32GB HBM memory. I'm reliably getting 60+ tokens per second, with the M2 Max trailing with around 45-50 and the Strix Halo last at around 35-40.

#LocalLLM

"Pays for itself". I'll rather pay for a #DGXSpark rather than hardware from four years ago.

https://www.theregister.com/systems/2026/05/21/amd-says-its-4k-ryzen-ai-halo-workstation-practically-pays-for-itself/5243755

#Technology #AMD #Radeon #StrixHalo #AI #LLM #Zen5 #RDNA3

Is anybody using a Strix Halo machine with 128GB uRAM?

I have a couple of questions...

#StrixHalo #localllm

My strix halo laptop has way better graphic performance when the internal display is on. With the display off the mouse cursor struggles to get 15fps. With the display on i can get 120Hz (on an external dispay). I don't want to keep using the OLED screen if I can avoid it. #Linux #amd #strixhalo

New year, new laptop! The #HP #ZBook Ultra G1a, based on AMD Strix Halo line of Ryzen AI CPUs, really shines with its 200+ GB/s memory bandwidth - where the Intel Core Ultra 9 mobile CPUs (2ch) only do ~100 GB/s. 128GB of quad-channel LPDDR5x-8533 FTW!

Also, I really like the new Linux Mint 22.3: the Cinnamon desktop looks great and hardware support for this particular ZBook is perfect!

#AMD #Ryzen #RyzenAI #StrixHalo @linuxmint

I think it's really kinda sad we haven't gotten a #Framework AMD #StrixHalo laptop before the RAM crisis happened :(

#AMD Just Showed Off Its Own DGX Spark Rival, The Ryzen #AIHalo: Full #ROCm Support, Day-0 Support For Leading #AI Models, Available In Q2
AMD Ryzen AI MAX family, codenamed #StrixHalo, has seen adoption, from laptops to handhelds and Mini PC; every consumer PC segment. High-performance and premium SoCs offer amazing performance thanks to #Zen5 #CPU, #RDNA 3.5 #GPU, & #XDNA 2 #NPU architectures. AMD developed its own Mini PC featuring these chips, called the #RyzenAIHalo.
https://wccftech.com/amd-showed-off-its-own-dgx-spark-rival-the-ryzen-ai-halo-full-rocm-support-q2-launch/

Working the Strix Halo in my ROG Flow z13. qwen3-vl:32b model, writing a report about why the sky is blue. 8.65 t/s. Same model wrote the ALT.

Now to fact check it (a fun hobby 😆)

#localai #llm #llms #qwen3 #strixhalo #amd #rog

(To the anti-AI people: please feel free to ignore/skip this post.)

New blog post! A few people have asked me how I run AI/LLM tools (ollama and llama.cpp) in my Talos Kubernetes cluster, so I decided to give it a proper explanation. This is quite a long and complex post, but I hope it's useful for someone.

#HomeLab #TalosLinux #Framework #ollama #llamacpp #AI #LLM #AMD #StrixHalo @homelab

https://mteixeira.wordpress.com/2025/12/12/running-ollama-and-llama-cpp-on-talos-linux-on-an-amd-strix-halo-cpu/

The article promises a deep dive into Strix Halo's memory subsystem and #iGPU challenges, but instead delivers the thrilling adventure of... enabling #JavaScript and cookies. 🍪🔧 Who knew that the real challenge was navigating modern web design? 📉🖥️
https://chipsandcheese.com/p/strix-halos-memory-subsystem-tackling #StrixHalo #MemorySubsystem #WebDesign #HackerNews #ngated

Strix Halo's Memory Subsystem: Tackling iGPU Challenges

https://chipsandcheese.com/p/strix-halos-memory-subsystem-tackling

#HackerNews #StrixHalo #MemorySubsystem #iGPUChallenges #TechNews #HardwareInnovation

Neue CPUs von AMD: Ryzen 9700X3D und AI Max+ 388/392 gesichtet (Update) https://www.computerbase.de/news/prozessoren/neue-cpus-von-amd-ryzen-9700x3d-und-ai-max-plus-388-392-gesichtet.94924/#update-2025-11-08T18:41 #ryzen #9700X3D #strixhalo

Neue CPUs von AMD: Ryzen 9700X3D und AI Max+ 388/392 gesichtet https://www.computerbase.de/news/prozessoren/neue-cpus-von-amd-ryzen-9700x3d-und-ai-max-plus-388-392-gesichtet.94924/ #ryzen #9700X3D #strixhalo

The OneXFly Apex is a handheld gaming PC with an AMD Strix Halo processor featuring discrete-class graphics, an 8 inch, FHD+ 120 Hz display, and an expected $1200(ish) starting price. Crowdfunding begins Nov 9: https://x.com/OnexPlayer_/status/1984183882707878147 #ONEXPLAYER #OneNetbook #OneXflyApex #HandheldGamingPC #StrixHalo

(To the anti-AI people: please feel free to ignore/skip this post.)

I may or may not write a blog post about setting up the Framework Desktop (AMD Ryzen AI Max+ 395) to run AI workloads with Talos Linux, but in case I don't, here is the TL;DR (too long; didn't read). Sorry, but needs to be a long Mastodon post, at least.

Start by going into the Talos Linux Image Factory to generate a system image that contains the AMD GPU drivers, then update (or install) the base system.

You can optionally install the Node Feature Discovery (NFD) by using it's standalone Helm chart. I chose that way because I want it to be independent from the AMD iGPU being present or not, but the next component can also install NFD for you. One trick I had to do, which I don't know whether it's required or not, was to create a custom NFD that sets "amd-gpu: true" when detecting vendor 1002, device 1586 (Strix Halo [Radeon Graphics / Radeon 8050S Graphics / Radeon 8060S Graphics] ).

Then you can use Helm charts to either install the AMD GPU Operator or the (simpler) AMD GPU Device Plugin for Kubernetes. I opted for the operator because of some of it's extra features, but it's up to you. The important thing is to disable the driver install, since this is handled natively by Talos itself. If you have installed NFD previously, don't forget to disable the install that comes with the operator.

Finally, you will install your AI/LLM workloads. I have both Ollama and Llama.cpp. The important thing here is that you need to hostPath mount /dev/dri and /dev/kfd in your containers. Ideally you should also pick one of the NFD labels as your node selector, just for good measure.

I hope this helps anyone. Feel free to ask any questions.

#HomeLab #TalosLinux #Framework #ollama #llamacpp #StrixHalo #AMD #Kubernetes @homelab

AMD's Chiplet APU: An Overview of Strix Halo: https://chipsandcheese.com/p/amds-chiplet-apu-an-overview-of-strix
#linux #hardware #amd #strixhalo #overview