# NousResearch/hermes-agent · GPU vendor detection and build selection fixes

## Open now

- 0 issues, 1 PR open on GitHub.
- PR #102505: feat(local-runtime): use SYCL builds on Intel GPUs + add vulkaninfo detection fallback

These members address incorrect GPU build selection in the local runtime, where AMD and Intel GPUs were not being matched to appropriate Vulkan or SYCL builds due to a vendor probe that only recognized NVIDIA hardware. Changes add proper vendor detection and a vulkaninfo fallback.

## TL;DR

- 4 member(s): 0 open issue(s), 1 closed issue(s), 1 open PR(s), 1 merged PR(s), 1 PR(s) closed without merging.
- A fix was merged to main in PR #131782. Verify it addresses your case.
- **Recommended:** read PR #131782 (and any competing PRs below) before filing. Confirm the merged change covers your symptom.

## Maintainer resolution verdict

- Internal-tracker verdict over this cluster's issue members: `unresolved`. This is distinct from the fix status above. Weigh it yourself.

## How this cluster was grouped

- Link evidence (`STRUCTURAL_SPANNING`): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.
- Pairwise link density `0.50`: the share of member pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36. `1.00` means every pair is linked.

## Closed / duplicate issues

- #100740: Managed local runtime installs CPU llama.cpp build on AMD/Intel GPU machines — vendor probe is NVIDIA-only, Vulkan never selected

## Prior & competing PRs

- **#131782: fix(local-runtime): AMD/Intel GPUs get the Vulkan build and are sized by their own memory (salvage #129750, #109336, #115076)**. merged to main · fix tier `verified`
  - maintainer direction: category `merged`
  - files: `contributors/emails/[redacted:email]`, `hermes_cli/local_runtime/bootstrap.py`, `hermes_cli/local_runtime/devices.py`, `hermes_cli/local_runtime/endpoint.py`, `hermes_cli/local_runtime/hardware.py`, `hermes_cli/web_routers/local_models.py`, `tests/hermes_cli/test_local_launch_budget.py`, `tests/hermes_cli/test_local_quickstart.py`, `tests/hermes_cli/test_local_runtime.py`, `tests/hermes_cli/test_unified_pool_quirk.py`, `website/docs/user-guide/local-models.md`
- **#100749: fix(local-runtime): use Vulkan builds on AMD and Intel GPUs**. closed without merge
  - maintainer direction: category `merged_via`; implication `superseded`; refs `8abd6dc2b5a`, `c8301ea6c9`, `#129750`, `#131782`
- **#102505: feat(local-runtime): use SYCL builds on Intel GPUs + add vulkaninfo detection fallback**. open
  - maintainer direction: category `uncategorized`; implication `needs_llm`
  - files: `.github/workflows/case-collision-check.yml`, `.github/workflows/ci-review-comment.yml`, `.github/workflows/ci.yaml`, `.github/workflows/contributor-check.yml`, `.github/workflows/deploy-site.yml`, `.github/workflows/docker-lint.yml`, `.github/workflows/docker.yml`, `.github/workflows/docs-site-checks.yml`, `.github/workflows/e2e-desktop.yml`, `.github/workflows/history-check.yml`, `.github/workflows/infographic-check.yml`, `.github/workflows/install-e2e-run.yml`, `.github/workflows/install-e2e.yml`, `.github/workflows/installer-tests.yml`, `.github/workflows/js-autofix.yml`, `.github/workflows/js-tests.yml`, `.github/workflows/label-rerun.yml`, `.github/workflows/lint.yml`, `.github/workflows/lockfile-diff.yml`, `.github/workflows/nix.yml` (+16 more)

## Fix status

- Merged to main in PR #131782 (cited above). No release data is tracked. Verify downstream.
- Fix last verified (merged) `2026-10-03`.
- Subsystem: `hermes_cli`