RT @Hi_MINISFORUM: Kann ein 351-Dollar-Mini-PC ein 30B+-KI-Modell lokal ausführen? 👀 Wir haben den MINISFORUM UM790 Pro mit Qwen 35B getestet — und die Ergebnisse könnten Sie überraschen. 💻 Ausgestattet mit: • AMD Ryzen™ 9 7940HS • Radeon™ 780M iGPU • 48GB DDR5 RAM (Dual-Channel) 🚀 Smooth-Inferenz mit 20+ Tokens/s (basierend auf unserem Test-Setup, Ergebnisse können variieren) Keine Cloud. Keine API-Gebühren. Nur lokale KI. 🔒 Learn more: s.minisforum.com/MiniPC #MINISFORUM #UM790Pro #LocalAI #MiniPC #AI #LLM #Tech Video Base Camp Bernie (@basecampbernie) Ein 300-Dollar-Mini-PC, der 26B-Parameter-Modelle lokal ausführt. Wir haben den MINISFORUM UM790 Pro mit AMD Radeon 780M iGPU, 48GB DDR5 RAM und 1TB NVMe getestet. Das Geheimnis: Die iGPU hat keinen dedizierten VRAM, sondern nutzt den Unified-Memory-Ansatz. Im BIOS sind nur 4GB VRAM angegeben, aber Vulkan erkennt den vollen Speicherpool. Wir haben 21+ GB für Modellgewichte zugewiesen. Er läuft: - Gemma 4 26B MoE: 19.5 tok/ a 20 tok/s - Gemma 4 E4B: 21.7 tok/s - Qwen3.5-35B-A3B: 20.8 tok/s - Dense 31B: 4 tok/s (Bandbreiten-Limit). Agentische Workflows mit @NousResearch Hermes Agent: Terminal, Dateioperationen, Web-Browsing, 40+ Tools, alles auf lokalen Modellen. Keine API-Keys. Nur eine Box auf dem Schreibtisch. Der RAM ist derzeit der Flaschenhals. DDR5-Preise sind 3-4x höher als letztes Jahr. Aber die Rechenpower ist nach dem Kauf forever free. @HiMINISFOR a-RDNA 3. Passt in die Hand. #LocalLLM #Gemma4 #llamacpp #AMD #Radeon780M #MoE #LocalAI #AI #OpenSource #GGUF #HermesAgent #NousResearch #DDR5 #MiniPC #EdgeAI #UnifiedMemory #Vulkan #iGPU…

mehr auf Arint.info

#agent #API #GGUF #llama #LocalAI #nitter #OpenSource #Qwen #Qwen3535 #arint_info

https://x.com/Hi_MINISFORUM/status/2046536248852885762#m

RT @basecampbernie: $300 mini PC running 26B parameter AI models at 20 tok/s. Minisforum UM790 Pro ($351) + AMD Radeon 780M iGPU + 48GB DDR5-5600 + 1TB NVMe. The secret: the 780M has no dedicated VRAM. It shares your DDR5 via unified memory. The BIOS says "4GB VRAM" but Vulkan sees the full pool. I'm allocating 21+ GB for model weights on a GPU with "4GB VRAM." The iGPU reads weights directly from system RAM at DDR5 bandwidth (~75 GB/s). MoE only activates 4B params per token = 2-4 GB of reads. That's why 20 tok/s works. What it runs: - Gemma 4 26B MoE: 19.5 tok/s, 110 tok/s prefill, 196K context - Gemma 4 E4B: 21.7 tok/s faster than some RTX setups - Qwen3.5-35B-A3B: 20.8 tok/s - Nemotron Cascade 2: 24.8 tok/s Dense 31B? 4 tok/s, reads all 18GB per token, bandwidth wall. MoE same quality? 20 tok/s. Full agentic workflows via @NousResearch Hermes agent with terminal, file ops, web, 40+ tools, all against local models. No API keys. Just a box on your desk. The RAM is the pain right now. DDR5 prices 3-4x what they were a year ago. But the compute is free forever after you buy it. @Hi_MINISFORUM @ggerganov llama.cpp + Vulkan + @UnslothAI GGUFs + @AMDRadeon RDNA 3. Fits in your hand. #LocalLLM #Gemma4 #llama_cpp #AMD #Radeon780M #MoE #LocalAI #AI #OpenSource #GGUF #HermesAgent #NousResearch #DDR5 #MiniPC #EdgeAI #UnifiedMemory #Vulkan #iGPU #RunItLocal #AIonDevice

Mehr auf Arint.info

#agent #API #GGUF #llama #LocalAI #OpenSource #Qwen3535 #arint_info

https://x.com/basecampbernie/status/2040326984446935059#m

Potenza e design in miniatura? Il Mini Incastro PRO con Ryzen 7 7840HS e grafica Radeon 780M integrata è qui per sorprendere! Silenzioso, elegante, perfetto per chi cerca un PC compatto senza compromessi.
Leggi la guida completa e scopri tutto su questo gioiellino
👉 https://incastro.altervista.org/mini-incastro-pro-potenza-e-raffinatezza-in-un-mini-pc/

#minipc #amd #ryzen7 #radeon780m #pcbarebone #gamingleggero #linux #tecnologia

So, good news. ROCm 6.3.4 and PyTorch 2.4.0 seems stable enough with gfx1103 if I use HSA override for 11.0.0, using latest firmware blobs and kernel 6.13.10 on Fedora 41.

In your Dockerfile, build your AI app from:
```
FROM rocm/pytorch:rocm6.3.4_ubuntu24.04_py3.12_pytorch_release_2.4.0
```

#rocm #pytorch #amdgpu #igpu #Radeon780M

Acer Nitro Blaze 11 e 8: handheld gaming con Ryzen 7
#Acer #CES2025 #GameNews #Gaming #Handheld #HandheldGaming #NitroBlaze #NitroBlaze11 #NitroBlaze8 #Notizie #Novità #Radeon780M #Ryzen7 #TechNews #Tecnologia

https://www.ceotech.it/acer-nitro-blaze-11-e-8-handheld-gaming-con-ryzen-7/

Snapdragon 8 Elite batte la Radeon 780M in 3DMark
#3DMark #AAA #Adreno830 #AMD #Benchmark #GameNews #Gaming #GamingMobile #GPU #Handheld #Hardware #Notizie #Novità #Prestazioni #Qualcomm #Radeon780M #ROGAlly #Smartphone #Snapdragon8Elite #Tecnologia

https://www.ceotech.it/snapdragon-8-elite-batte-la-radeon-780m-in-3dmark/

AMD Ryzen 7 8700G Review: Zen 4 For The Budget-Minded #8700g #amd #amdryzen #amdryzen78700g #features #hardware #processor #radeon780m #reviews #ryzen78700g #ryzenai

https://www.lowyat.net/2024/316637/amd-ryzen-7-8700g-review/

🕹️ Orange Pi Neo un palmare gaming con AMD Ryzen e Manjaro Linux
Orange Pi Neo un palmare Linux con sistema operativo Manjaro Gaming Edition e hardware di fascia alta.

https://gomoot.com/orange-pi-neo-un-palmare-gaming-con-amd-ryzen-e-manjaro-linux

#amd #bluetooth #GaN #Hall #lcd #Linux #manjaro #arch #OLED #orangepi #orangepineo #radeon780m #zen4 #ryzen #steam #steamdeck #wifi #vrr

Arriva un nuovo contendente per la Steam Deck e questa volta parla Linux ! I palmari sembrano essere una vera tendenza in questo momento dopo che Steam Deck ha iniziato le danze. Asus ROG Ally, Legion Go, MSI Claw tutti con OS Windows. Ed ora viene annunciato l’ Orange Pi Neo con Manjaro Linux, un palmare gaming compatibile con Steam.

Perchè la scelta […]

https://gomoot.com/orange-pi-neo-un-palmare-gaming-con-amd-ryzen-e-manjaro-linux/