docs: tested configurations: RTX 4070, 615.71, Ubuntu 26.04 GNOME (v2.1.0-rc.3 nightly, #2314) #2428

Merged

jabbaxipil merged 1 commit from docs/tested-config-rtx4070 into main

2026-11-12 03:04:35 +00:00

jabbaxipil commented

2026-11-11 18:54:42 +00:00

(Migrated from github.com)

Adds four rows to docs/tested-configurations.md for a plain RTX 4070 on Linux with driver 615.71, GNOME Wayland and a 5800X3D on Linux. The page had none of these yet.

The rc.3 nightly at a fixed Kokiri view: 59.2 fps, CPU-limited.
The same build with #2314's vk_memory.cpp, with and without RX_VK_SHADOW_STAGE=0: similar on a fresh launch (59.0 and 63.2 fps); after about 10 minutes of exploring, 49.5 fps with staging and 33.1 without.
The rc.3 nightly with frame generation 3x: 60.9 fps at the same Kokiri spot, then a device lost (Xid 109) in Hyrule Field about 8 minutes in.

See #2314 (the measurements there in full), #1000 (the device lost, with its log).

Assisted-by: Claude Opus 5.5 (measurements, analysis, drafting)

🤖 Generated with Claude Code

Adds four rows to `docs/tested-configurations.md` for a plain RTX 4070 on Linux with driver 615.71, GNOME Wayland and a 5800X3D on Linux. The page had none of these yet. - The rc.3 nightly at a fixed Kokiri view: 59.2 fps, CPU-limited. - The same build with #2314's `vk_memory.cpp`, with and without `RX_VK_SHADOW_STAGE=0`: similar on a fresh launch (59.0 and 63.2 fps); after about 10 minutes of exploring, 49.5 fps with staging and 33.1 without. - The rc.3 nightly with frame generation 3x: 60.9 fps at the same Kokiri spot, then a device lost (Xid 109) in Hyrule Field about 8 minutes in. See #2314 (the measurements there in full), #1000 (the device lost, with its log). Assisted-by: Claude Opus 5.5 (measurements, analysis, drafting) 🤖 Generated with [Claude Code](https://claude.com/claude-code)

Joey-Magnus (Migrated from github.com) reviewed

2026-11-11 20:08:39 +00:00

Joey-Magnus (Migrated from github.com) left a comment

should fix — conflict: docs/tested-configurations.md conflicts with main, because other rows have landed at the top since. Rebase and keep your three rows at the top, newest first.

minor — one run per row: the third row (docs/tested-configurations.md:19 at this head) ends with "An earlier session with frame generation 3x lost the device once (Xid 109, see #1000)". That's a different run with different settings (the row's settings say frame generation off), and the page is one row per run. Please either give it a row of its own with its build, settings and spot, or move it to #1000 and drop it here. Since it's a device lost on a configuration the page didn't have, it's worth its own row if you have the details.

Thanks, this fills real gaps. The page had no plain RTX 4070, no 615.71 driver and no GNOME session until now (only the 4070 Ti and 4070 SUPER). The rows give what the page asks for: the build string from the log, hardware, settings, a named spot, and fps from the present lines. I checked the #2314 details against that PR: its head is `97c1a202`, and `RX_VK_SHADOW_STAGE=0` is the switch it adds, so the first two rows describe a build others can reproduce. **should fix — conflict:** `docs/tested-configurations.md` conflicts with main, because other rows have landed at the top since. Rebase and keep your three rows at the top, newest first. **minor — one run per row:** the third row (`docs/tested-configurations.md:19` at this head) ends with "An earlier session with frame generation 3x lost the device once (Xid 109, see #1000)". That's a different run with different settings (the row's settings say frame generation off), and the page is one row per run. Please either give it a row of its own with its build, settings and spot, or move it to #1000 and drop it here. Since it's a device lost on a configuration the page didn't have, it's worth its own row if you have the details.

jabbaxipil commented

2026-11-11 20:34:31 +00:00

(Migrated from github.com)

Thanks. Rebased onto main with our rows still at the top. The Xid 109 session now has its own row (frame generation 3x, its own Kokiri fps, and where the device was lost), and the sentence is gone from the third row.

Joey-Magnus (Migrated from github.com) approved these changes

2026-11-11 21:22:30 +00:00

Joey-Magnus (Migrated from github.com) left a comment

Rebased, the rows are at the top, and the Xid 109 session has its own row. Each row has one run with its build, settings, spot and fps. Good to go.

hckrman101 commented

2026-11-11 22:55:52 +00:00

(Migrated from github.com)

This now conflicts with main in docs/tested-configurations.md. Could you merge main into the branch (or rebase) when you get a chance? I'll take another look once it's up to date.

This now conflicts with main in `docs/tested-configurations.md`. Could you merge main into the branch (or rebase) when you get a chance? I'll take another look once it's up to date.

jabbaxipil commented

2026-11-12 02:19:44 +00:00

(Migrated from github.com)