About this site

GPU Unlocker is an independent guide to an open-source project we didn't write — built because the project's results are striking, its documentation is scattered across READMEs and issue threads, and people deserve the risks in the same font size as the benchmarks.

What we do

Explain — how the gpu-unlock kernel module works, in plain order, following the project's own documentation.
Gather — the published benchmark figures and compatibility reports, in one place, each labelled by origin.
Document the edges — the known exception card, the workloads that don't speed up, the failure mode, the missing gaming data. The caveats are content here, not footnotes.

Our sourcing rules

Every performance figure on this site traces to the upstream projects' documentation or the reports they link: Zagitta/gpu-unlock and JustAResearcher/gpu-unlock-ada.
Figures are labelled project-reported (the maintainer's published measurements) or community-reported (user reports the project collects). We have not independently re-run them, and we say so.
Where the documentation gives no number, we print no number. If a claim can't be sourced — a percentage without a methodology — it doesn't go on the site.
Pages carry a "checked" date. If upstream results change, we update and re-date.

What we are not

Not the gpu-unlock project, and not affiliated with its authors.
Not affiliated with, or endorsed by, NVIDIA Corporation. GeForce, RTX and NVIDIA are NVIDIA's trademarks.
Not a support channel for the module — installation problems belong in the upstream issue trackers, where the people who can fix them actually read.
Not a promise. This site documents reported results; whether to modify how your hardware boots is your decision and your risk.

Corrections

Found an error, a stale figure, or a card result we've missed? Tell us — corrections that come with a source get applied and re-dated.

Who runs this site

GPU Unlocker is maintained independently of the projects it covers — no sponsorships, no vendor relationships, and no affiliate deals with NVIDIA or the projects' authors. The maintainer is reachable through the contact page; corrections are applied with a source and logged below.

Corrections & updates

2026-10-25 — Site launched. All figures pinned to the upstream READMEs (gpu-unlock commit af28cbb; gpu-unlock-ada commit 4e7fefc).
2026-10-25 — Post-launch independent dual review (ChatGPT + Gemini): claim wording tightened to match the underlying reports; a fully-attributed ComfyUI community result added to the benchmarks page; the compatibility quick-check gained the OEM exception entry, a PCI-ID lookup and per-status next steps; mobile navigation rebuilt as a proper menu.