diff options
Diffstat (limited to 'assets/2026-08-13-velox-amd-board-swap-context.org')
| -rw-r--r-- | assets/2026-08-13-velox-amd-board-swap-context.org | 5 |
1 files changed, 5 insertions, 0 deletions
diff --git a/assets/2026-08-13-velox-amd-board-swap-context.org b/assets/2026-08-13-velox-amd-board-swap-context.org new file mode 100644 index 0000000..531c365 --- /dev/null +++ b/assets/2026-08-13-velox-amd-board-swap-context.org @@ -0,0 +1,5 @@ +#+TITLE: Velox context for the microcode check and possible reinstall +#+SOURCE: from work +#+DATE: 2026-08-13 14:34:48 -0500 + +Velox context for the microcode check and possible reinstall. Velox's mainboard was swapped today from Intel to AMD (Ryzen AI 9 HX 370, Radeon 890M, 96GB new RAM, trained fine). The old SSD is intact but the new board's NVRAM is empty, so there is no boot entry: the fix in progress from the work session is boot the Arch USB (Secure Boot enforcement must be disabled first, factory default is on), chroot, efibootmgr --create --disk /dev/nvme0n1 --part 1 --label GRUB --loader '\EFI\GRUB\grubx64.efi', then swap intel-ucode for amd-ucode, grub-mkconfig, mkinitcpio -P. Craig may instead do a full reinstall via archsetup. The ask he is bringing to this session: verify archsetup installs the correct microcode on AMD — whether it hardcodes intel-ucode or detects the CPU vendor — and fix it if it is Intel-only, since velox is now AMD and flies with him Sunday. Also note: the new board reset the RTC to 2025-01-01, so first boot may hit TLS/pacman signature grumpiness until the clock syncs. |