Arch Linux first-run setup

Two tracks

Zero to first token

1. 1 · Cited, not verified on our rig

   Rolling release and kernel notes

   AMD’s 10.1 matrix does not list Arch. Arch is rolling — there is no “Arch 24.04” row to pin. Prefer the stock linux kernel and the in-tree amdgpu module Arch Wiki AMDGPU documents. A custom kernel, linux-zen experiment, or DKMS amdgpu is not verified on our rig. Dual-boot Ubuntu 24.04.5 if you want the measured local desktop path.

   • Ubuntu setup (measured)
   • R9700 measured page
   • GPU directory

2. 2 · Cited, not verified on our rig

   render and video groups

   AMD Linux prerequisites add the login to render and video: sudo usermod -aG render,video $USER. Arch Wiki users and groups apply the same idea. Log out, or reboot, before you trust rocminfo or /dev/kfd. Root “works” by skipping the check — that is not the fix. Not verified on our rig.

   • Verify ROCm (Ubuntu)
   • What is ROCm

3. 3 · Cited, not verified on our rig

   Mesa RADV Vulkan first

   Consumer first token on AMDai is llama.cpp + Vulkan. Arch Wiki AMDGPU and Vulkan: install mesa, vulkan-radeon, vulkan-icd-loader, and vulkan-tools, then vulkaninfo. Look for the discrete Radeon and RADV — not AMDVLK. If vulkaninfo is empty, stop; do not guess a driver. BIOS notes (Above 4G, ReBAR/SAM, CSM off) apply on the same board. Not verified on our rig.

   • llama.cpp
   • Vulkan vs HIP
   • BIOS

4. 4 · Cited, not verified on our rig

   ROCm / HIP from Arch packages

   After Vulkan tokens are the goal, Arch Wiki GPGPU documents the extra ROCm packages (HIP / OpenCL on the amdgpu kernel driver). That is Arch’s build, not AMD Core SDK 10.1. Confirm rocminfo prints the gfx the card actually is. Arch Wiki notes ROC_ENABLE_PRE_VEGA=1 for Polaris-era cards — do not set that on gfx1201, and do not invent an HSA override. Rolling upgrades of that extra set are a distro story — see /rocm/upgrade. Not verified on our rig.

   • Install ROCm (Ubuntu)
   • Upgrade or reinstall ROCm
   • ROCm problems

5. 5 · Cited, not verified on our rig

   Verify: vulkaninfo, rocminfo, then a monitor

   Order: vulkaninfo (RADV + discrete Radeon), then rocminfo if you installed Arch ROCm packages. A userspace monitor such as amdgpu_top (Arch extra) can show the card busy after tokens start — it is not required for first token and is not verified on our rig. We do not invent an Arch tok/s.

   • Verify ROCm (Ubuntu)
   • Troubleshoot

6. 6 · Cited, not verified on our rig

   llama.cpp — Vulkan first, then HIP

   Arch Wiki llama.cpp: install llama-cpp plus ggml-vulkan first. HIP is second: ggml-hip only after rocminfo lists this gfx. Or build from ggml-org with the flags build.md documents: cmake -B build-vk -DGGML_VULKAN=ON. Ollama, Strata, and vLLM are not an Arch first-run on AMDai. Not verified on our rig.

   • llama.cpp
   • Backends
   • Speed up llama.cpp

7. 7 · Cited, not verified on our rig

   Pick Q6, Q5, or Q4 — Q3 is a labeled estimate

   Download the same official-instruct GGUF the Ubuntu recipes name. Start at Q6, Q5, or Q4 and check VRAM + KV on the calculator. Q3 stays a labeled estimate unless an Arch run card exists — none does in this repo. Q1/Q2 stay off. Tok/s on /benchmarks and /results are Ubuntu llama-bench on the local desktop. Treat them as comparison, not an Arch promise.

   • Calculator
   • 35B recipe
   • Quants
   • Qwen compare

8. 8 · Cited, not verified on our rig

   Fit VRAM and the two RAM lines

   The calculator prints Minimum system RAM needed (the load at the context you picked) and RAM needed at full context (262144 for the house 35B). Overnight UD-Q5 one-GPU Ubuntu peaks are 37.32 GB at 4K and 42.9 GB at 256K — measured on the local desktop, not an Arch kit. Keep the GGUF on NVMe. Not verified on our rig.

   • Calculator
   • Flash RAM offload
   • Benchmarks

9. 9 · Cited, not verified on our rig

   First-run command (Ubuntu-measured shape)

   Use the house daily-driver launch from the 35B recipe / tuning guide: ./build-vk/bin/llama-server -m ~/models/Qwen3.6-35B-A3B-UD-Q5_K_XL.gguf -ngl 99 -c 262144 --cache-type-k f16 --cache-type-v f16 --host 127.0.0.1 --port 8080. That command is the Ubuntu local desktop shape. On Arch it is not verified on our rig and not measured on the local desktop — confirm vulkaninfo, then tokens, then a loopback /v1. Do not copy Ubuntu tok/s onto this box. Read the 4K run card to see what “good” looked like on the measured path.

   • 35B recipe
   • 4K run card
   • Run a local Qwen
   • Tuning flags

10. 10 · Cited, not verified on our rig

    When it breaks

    Start from a symptom on /troubleshoot. Paste a log in the error matcher. Cited rows live on /rocm-problems. Firmware: /bios. Jargon: /glossary. Short answers: /faq. Stack Lab remains the Ubuntu companion — packages are not published yet and there is no Arch package. Not verified on our rig.

    • Troubleshoot
    • Error paste
    • ROCm problems
    • FAQ
    • Glossary
    • Stack Lab

RAM lines (Ubuntu-measured)

Use the calculator before you download a second quant. The two labels stay Minimum system RAM needed and RAM needed at full context. The house 35B peaks below are Ubuntu, not Arch.

Open in Stack Lab

Stack Lab remains Ubuntu-first. The button is the allowlisted scheme. There is no Arch package and no public download.

Sources

• AMD — ROCm 10.1.0 compatibility matrix (official OS rows)
• AMD — Install ROCm 10.1.0 (apt / dnf / zypper on listed OS rows)
• AMD — Linux prerequisites (render and video groups)
• Arch Wiki — AMDGPU (mesa, vulkan-radeon)
• Arch Wiki — Vulkan (vulkan-icd-loader, vulkan-radeon)
• Arch Wiki — GPGPU / ROCm
• Arch Wiki — llama.cpp (ggml-vulkan, then ggml-hip)
• llama.cpp — build.md (Vulkan / HIP)