[Tracking] AMD rendering: white silhouettes, block squares / grid lines, no anti-aliasing (networks in software)

This issue collects every open report of AMD-specific rendering problems, so there is one place to follow it. The individual reports were closed pointing here; their logs, screenshots and discussion stay in each linked issue.

Summary

AMD-specific picture problems on Windows:

Link or buildings drawn as flat white silhouettes;
black-outlined squares or blocky textures;
dark grid lines on Hyrule Field's terrain;
no anti-aliasing or a softer picture, because the game's image networks can't run fast on AMD;
white geysers;
glowing tunic trims.

Affected GPUs and versions

GPUs: RX 6700 / 6700 XT / 6750 XT, RX 7900 XTX, RX 9060 XT / 9070 XT, Radeon 860M / 890M.
Versions: early builds to v2.0.5.

Evidence from the logs

graphical bugs amd #224 (10 people):
- black-outlined squares, white sky, missing meshes, fog;
- on the 890M, Link drawn as a flat white silhouette (rejected computes 0775215f / 72ba090c / f10782f9 on 171x96);
- 2-minute freezes while compute pipelines compile.
Missing and inaccurate textures, blurry/pixelated text and map (AMD, log attached) #57: blocky squares from the ASTC-to-BC3 re-encode on AMD; RX_VK_ASTC_RGBA8=1 works around it.
v2.0.5's terrain re-encode (vk: BC3 levels that halve evenly on AMD (the terrain grid in Hyrule Field) #705): caused the rainbow ground and 70-140 ms GPU time. v2.0.6 turned it off; the release notes say RX_VK_UNEVEN_BC=0 turns it on for testing, and the thin grid lines on Hyrule Field may come back on AMD until it is redone.
v2.0.6 also includes PR BC3 re-encode of ASTC arrays: one layer a run (Hyrule Field grid lines) #956: the BC3 re-encode of ASTC arrays now does one layer a run, which removed the dark tile outlines on Hyrule Field's terrain at a distance on an RX 6600. RX_VK_BC_LAYERS=<n> restores multi-layer runs.
AMD RX 9070 XT: OoT's networks fall back to software matrices (no 16x8x16 cooperative matrix), ~56 ms GPU per frame #52 / Vulkan Compute Shader Translation Failure #628: the TAAU / upscale networks run in software on RDNA3/4 (no 16x8x16 cooperative matrix), so they are left out on non-NVIDIA GPUs: a stretched picture and no AA. The proposal is 16x16x16 through shared memory. FSR 4 is an open PR (Add opt-in FSR 4 Native AA on Windows Vulkan #577).
AMD feedback (RX 6750 XT) and a couple of suggestions #305: bright tunic trims, shimmering.
[AMD] Dodongo's Cavern shader issues #769: Dodongo's Cavern geysers drawn fully white.
New feedback in the last version on the specs: Ryzen 5 5600X & AMD Radeon 6700XT #596: castle and town drawn as white silhouettes at first.

Workarounds

Blocky squares: RX_VK_ASTC_RGBA8=1 (uses more VRAM).

Reports merged here (9 issues, about 25 people)

Issue	Title	People	GPU	Version	Key detail
#224	graphical bugs amd	10	RX 6700/6700 XT, Radeon 890M, RX 9070 XT	v2.0.2	AMD tracker: black outlined squares, white sky, missing meshes, fog. 890M: v2.0.2 fixed device lost (bounded CAS); left: Link drawn flat white silhouette (rejected compu…
#305	AMD feedback (RX 6750 XT) and a couple of suggestions	4	RX 6750 XT	v2.0.2	AMD feedback: bright/glowing tunic trims, shimmering, no AA (requests: >100% res, SMAA/FXAA). v2.0.2: "Preparing shaders" hangs at 99% on AMD from graphics pipeline 5fa5…
#52	AMD RX 9070 XT: OoT's networks fall back to software matrices (no 16x…	2	RX 9070 XT, RX 7900 XTX (RADV)	important+	Networks run in software on RDNA3/4 (no 16x8x16 coopmat); proposal: 16x16x16 via shared memory; networks now left out on non-NVIDIA (stretched picture, no AA)
#57	Missing and inaccurate textures, blurry/pixelated text and map (AMD,…	2	RX 7900 XTX	early	Blocky squares = ASTC->BC3 re-encode on AMD; workaround RX_VK_ASTC_RGBA8=1; #45 (closed dup) reporter says squares still there; texture corruption fix in faster-fps
#596	New feedback in the last version on the specs: Ryzen 5 5600X & AMD Ra…	2	RX 6700 XT	v2.0.2	RX 6700 XT: Link over-bright in cutscene, texture flicker in Kokiri, stuck on last 2 shaders at restart, AMD driver crash (black no-signal, reboot) in Kokiri on 26.10.1;…
#628	Vulkan Compute Shader Translation Failure	2	Radeon 860M (Win + RADV)	honest-eta	Networks left out on RDNA3.5 iGPU (no AA at 100%, #52); Windows: fails to load after the owl; Linux: 21.5 fps CPU/submit-bound (#528)
#197	Game now launches on RX 9070 XT with the latest build, but textures/c…	1	RX 9070 XT	pre-foliage	Wrong colours (#224), stutter, VRAM; crash entering Deku Tree (read guest 0x8)
#769	[AMD] Dodongo's Cavern shader issues	1	RX 9060 XT	v2.0.3	Dodongo's Cavern geysers render full white (AMD)
#1017	RX 6700 XT: 10-11 shader programs fail to translate from the title sc…	1	RX 6700 XT (driver 32.0.21045.16010)	v2.0.5	10 programs fail to translate from the title screen, 11 in Hyrule Field (frame-stat counter only, no names in the log); asks for a flag that names them

Adding a report

Please update to v2.0.6 first (https://github.com/NotRexed/Hookshot/releases/tag/v2.0.6; Update next to Play in the launcher). Please add: GPU, driver version, screenshot, version, and hookshot_log.txt.

All tracking issues

[Tracking] VRAM and RAM fill up over play time: fps fades, textures degrade or turn rainbow, a restart fixes it (mostly 4-8 GB cards) #993 Garbled / rainbow textures that get worse over play time (VRAM over budget, mostly 4-8 GB cards)
[Tracking] Floating trees, missing mountains and vanishing near vegetation #994 Floating trees, missing mountains and vanishing near vegetation
[Tracking] Low FPS with low GPU usage (CPU-bound), FPS drops in Castle Town / Hyrule Field, shader stutter #995 Low FPS with low GPU usage (CPU-bound), FPS drops in Castle Town / Hyrule Field, shader stutter
[Tracking] AMD on Windows: driver crash, device lost or freeze (RX 5000-9000, Deku Tree / after cutscenes) #996 AMD on Windows: driver crash, device lost or freeze (RX 5000-9000, Deku Tree / after cutscenes)
[Tracking] AMD handhelds and APUs (ROG Ally / Ally X, Legion Go, Radeon 780M/860M/890M): black screen, device lost, reboot #997 AMD handhelds and APUs (ROG Ally / Ally X, Legion Go, Radeon 780M/860M/890M): black screen, device lost, reboot
[Tracking] Black screen with audio after the intro / at the Deku Tree cutscene #998 Black screen with audio after the intro / at the Deku Tree cutscene
[Tracking] Intel Arc and Core Ultra iGPUs: device lost, black screen after the intro, Linux ANV crashes #999 Intel Arc and Core Ultra iGPUs: device lost, black screen after the intro, Linux ANV crashes
[Tracking] Linux and Steam Deck: GPU hangs, crashes, low FPS and build failures #1000 Linux and Steam Deck: GPU hangs, crashes, low FPS and build failures
[Tracking] GPUs without mesh shaders (GTX 9xx/10xx, RX 400/500/5000, Iris Xe/UHD): v2.0.6 fallback reports #1001 GPUs without mesh shaders (GTX 9xx/10xx, RX 400/500/5000, Iris Xe/UHD): v2.0.6 fallback reports
[Tracking] Invisible or missing objects, walls and effects #1002 Invisible or missing objects, walls and effects
[Tracking] Antivirus false positives on HookshotBuilder.exe (Defender Trojan:Script/Wacatac.H!ml) #1003 Antivirus false positives on HookshotBuilder.exe (Defender Trojan:Script/Wacatac.H!ml)
[Tracking] Lighting: overexposure, over-bright bloom, crushed or black shadows, black faces #1004 Lighting: overexposure, over-bright bloom, crushed or black shadows, black faces
[Tracking] Flicker and flashing: brightness pulses, one-frame flashes, doubled image with frame pacing #1005 Flicker and flashing: brightness pulses, one-frame flashes, doubled image with frame pacing
[Tracking] AMD rendering: white silhouettes, block squares / grid lines, no anti-aliasing (networks in software) #1006 AMD rendering: white silhouettes, block squares / grid lines, no anti-aliasing (networks in software)
Feature requests (collected) #1007 Feature requests (collected)

Full hookshot_log.txt from v2.0.6 on an RX 6700 XT as requested in #1017: https://gist.github.com/kafasam/87d289c0f48d93eb5ed45a3431cf9ef8

Context: 10 programs failed at the title, 11 in Hyrule Field; the same session ran 4K output (HOS_OUTPUT_RES=2) + Resolution 75% + FG2, with visible symptoms of textures not loading / arriving late on area re-entry (Kokiri after Hyrule Field hung a long while before textures filled in), ~22 fps real in Hyrule Field, failed-programs count unchanged from v2.0.5. Happy to run any instrumented build or capture frames.

Follow-up from the same RX 6700 XT (v2.0.6, 4K output / Resolution 75% / FG2): two log patterns that look like the texture side of this tracker.

1. Slow ASTC(BC3) uploads stalling frames during play — streaming a new texture costs 21–27 ms each on the upload thread, e.g. in Hyrule Field (and stacking on area re-entry, which we see as multi-second hangs while textures fill in):

vk: slow upload 27.2 ms: #9882 0x1140339000 683x683x1 x1 levels 10 NVN 0x8c ASTC (BC3), 271 of 271 KB
vk: slow upload 25.9 ms: #9885 0x10e8b43e00 683x683x1 x1 levels 10 NVN 0x8c ASTC (BC3), 271 of 271 KB
vk: slow upload 21.7 ms: #9888 0x10e8f3ee00 683x683x1 x1 levels 10 NVN 0x8c ASTC (BC3), 271 of 271 KB

2. The import pool never releases: imports: 81 live (2208 MB), 63 retired still imported (1840 MB), 0 retired given back (0 MB) — 1243 live textures and climbing across a session. Retired memory is never returned, so the game keeps importing/re-uploading rather than reusing.

Given the failed compute programs on AMD are also texture-side (our counts in #1017), we're testing RX_VK_ASTC_RGBA8=1 next to bypass the BC3 re-encode entirely — will report whether the slow uploads and the streaming freezes disappear.

RX 6700 XT follow-up on the RX_VK_ASTC_RGBA8=1 test we said we'd run: it bypasses the BC3 re-encode but is not viable at this game's texture volume on a 12 GB card.

Live numbers from Hyrule Field/Lon Lon Ranch, same session shape as the earlier reports (4K output, Resolution 75%, FG2):

	default (BC3)	RX_VK_ASTC_RGBA8=1
texture memory	2,774 MB	6,612 MB and climbing
GPU time/frame	21–28 ms	86–90 ms (3–4×)
real fps	19–26	10.6–11.2

So the 16–27 ms ASTC(BC3) slow uploads stand as the least-bad path here; the streaming stalls and area-entry hangs remain (reproduced at Lon Lon Ranch on v2.0.6 with the default path). The failed-compute-programs connection in this tracker remains our best explanation for why the re-encode is slow on AMD while the same content streams fine elsewhere — everything still points at the texture-side compute kernels.

AMD testers: green diagonal stripes on v2.0.6+ (RX 7600 report on Discord)

v2.0.6 turned on two AMD-only shader changes. If you see bright green diagonal stripes on trees, terrain or characters, please try each of these and reply with what changed (one at a time, same spot):

Make a file test.bat in your game folder (next to hookshot.exe) with:

@echo off
set RX_VK_NO_INVARIANT=1
hookshot.exe

Then the same with set RX_VK_NO_NAN_GUARD=1 instead. Please include your GPU, driver version and hookshot_log.txt. Thanks!

crashing also on 6800xt, freezing and closing the app.

RDNA2 report for the stall side of this tracker, v2.0.7 (source 9b541c9c), Windows 11, RX 6900 XT 16 GB (AMD proprietary 2.0.353 / Adrenalin 26.10.1), i7-9700K, 32 GB. Saves moved over from another build; Hyrule Field, then Castle Town and back out.

Picture: no green blobs or block squares on this card. Between stalls 38–41 fps (vi: ... fps over the last 5.0 s). Terrain holes where the sky shows through the hills in Hyrule Field, which is #994. The image networks are left out as expected (not translated: a network left out on this card).

Stalls: single submits of many seconds, nearly all of it in draw:

vk: slow submit: 43129.3 ms: draw 43118.3 descriptors 125.0 texture lookups 128.2 uploads 52.8 pipelines 2.1 programs 38.9 commands 10.6
vk: slow submit: 41337.7 ms: draw 41337.6 descriptors 7.9 texture lookups 5.7 uploads 4.2 programs 2.6 commands 1.9
vk: slow submit: 13355.5 ms: draw 13355.0 ...
vk: slow submit: 13109.0 ms: draw 13108.8
vk: slow submit: 11512.7 ms: draw 11512.6
vi: 3930 frames presented (5.0 fps over the last 48.0 s, interval 1)

79 slow submits in that session. Leaving a house, breaking pots and opening the menu each brought one on. The lines right before the 43 s one were the terrain material arrays being decoded and re-encoded, with the texture budget at its cap:

vk: texture 0x109538a600 1365x1365x1 x90 levels 11 NVN 0x8e -> vk 137 (ASTC decoded) (2206 textures, 5712 MB)
vk: texture 0x1098c02c00 1365x1365x1 x90 levels 11 NVN 0x80 -> vk 137 (ASTC decoded) (2207 textures, 5930 MB)
...
vk: VRAM: 10779 MB used of 13138 MB allowed (1886 MB of it free in the blocks); textures 6137 MB (budget 6144 MB), 2972 textures

So the textures sat at the 6 GB cap from vk_texture.cpp (min(allowed * 0.7, 6 GB), the comment about 16 GB cards in #105/#106), with ~2.3 GB of the driver's allowance still unused. My guess was eviction and re-decode churn of those 90-layer arrays, so I tried:

Run	Result
default	12.8 s stall in Castle Town, then 13.8 fps over 17 s
`RX_VK_NOOPT=1`	stalls of 11–43 s (the list above), 38–41 fps between
`RX_VK_NOOPT=1` + `RX_VK_VRAM_MB=8192`	still stalls by the player's account; I have no numbers for this run

So raising the budget didn't remove them, and I'm reporting this as data rather than a cause. What stands out is that the time is booked under draw with pipelines near zero, so it isn't the driver building pipelines (RX_VK_NOOPT is set). The build is gone from this PC now, so I can't capture more, but the numbers may help whoever looks at the AMD stalls.

More reports of this, collected here:

Frayed/fringed cloth renders solid white or black (Kokiri clothes, pots, Sheik's wrap, Talon's tassels) on RX 6700 XT #1540: Frayed/fringed cloth renders solid white or black (Kokiri clothes, pots, Sheik's wrap, Talon's tassels) on RX (v2.0.11; AMD Radeon RX 6700 XT, Vulkan 1.4, Windows, 2560x1440; Ryzen 5 7500F, Windows 11 Home)
[AMD RX 6700 XT] Rainbow / multicoloured glowing blocks on crystals in Dodongo's Cavern since v2.0.12 (fine on v2.0.11): BC6H/BC5 raw colour attachment fails with -11 #1801: [AMD RX 6700 XT] Rainbow / multicoloured glowing blocks on crystals in Dodongo's Cavern since v2.0.12 (fine on (v2.0.13; AMD Radeon RX 6700 XT (12 GB), Adrenalin [26.10.x] (log: Vulkan driver; Ryzen 5 5600X, 32 GB RAM, Windows [10/11])
[AMD 6650 XT] Textures replaced with a glowing color #2054: [AMD 6650 XT] Textures replaced with a glowing color (v2.0.14; Radeon RX 6650 XT, 32.0.21030.2001; Ryzen 5 3600, 32GB RAM, Windows 10)
Network left out on this card issue #2136: Network left out on this card issue
Heavy visual glitches #2149: Heavy visual glitches (v2.1.0; rx 6700, 26.10.2; Ryzen 7 5700x3d, 32gb ram, Windows 26h2)
Link Glowing White During Opening Cutscene AMD 7900xt GPU #2191: Link Glowing White During Opening Cutscene AMD 7900xt GPU (v2.1.0; 7900XT 26.9.1; Ryzen 9 9900x, 16GB DDR5, Windows 11)
Glowing Textures on AMD 6600xt #2259: Glowing Textures on AMD 6600xt
AMD RX 6750 XT, 2.1.0: white line artifacts (grass and mountain outline) and low fps in open areas #2444: AMD RX 6750 XT, 2.1.0: white line artifacts (grass and mountain outline) and low fps in open areas (v2.1.0; RX 6750 XT 12 GB (Adrenalin 26.10.1); Intel i7-11700F, 32 GB RAM, Windows 11)

More reports of this, collected here:

Compute shader translation failure (program not translated: a network left out on this card) #2569: Compute shader translation failure (program not translated: a network left out on this card) (v2.1.0; RX 6600 8GB, 26.10.1; Ryzen 5 4600g, 16 GB (2x8), Windows 11 26h2)

Same on XFX AMD Radeon 6700XT running Windows 10 LTSC 2021 IoT

is it fixed?

Tracker upkeep note (merged pull requests and code checked against main 51da0464; no new test run was made for it).

The older summary's ‘no anti-aliasing’ description is now too broad: main includes #1115's FXAA stand-in, #1552's guarded RGB9E5 gameplay fallback and #2166's fixed-resolution FSR path, alongside several subgroup and cloth fixes. Full native matrix-network support is still not present for every AMD device: main 51da0464 selects 16×8×16 f16 hardware and otherwise retains network admission limits and stand-ins. #1975 owns staged-matrix support, with separate BC-attachment and RADV-shuffle work in #2001 and #1893. Please mark the completed subcases but keep this tracker open; a ‘network left out’ message alone does not diagnose a texture-decoding failure.

is it fixed?

Wasn't as of last night

Navigation Menu

Repository navigation

[Tracking] AMD rendering: white silhouettes, block squares / grid lines, no anti-aliasing (networks in software) #1006

Summary

Affected GPUs and versions

Evidence from the logs

Workarounds

Reports merged here (9 issues, about 25 people)

Adding a report

All tracking issues

12 remaining items

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

[Tracking] AMD rendering: white silhouettes, block squares / grid lines, no anti-aliasing (networks in software) #1006

Description

Summary

Affected GPUs and versions

Evidence from the logs

Workarounds

Reports merged here (9 issues, about 25 people)

Adding a report

All tracking issues

Activity

kafasam commented on Nov 9, 2026

kafasam commented on Nov 9, 2026

kafasam commented on Nov 9, 2026

NotRexed commented on Nov 9, 2026

renanschu-png commented on Nov 9, 2026

lev92 commented on Nov 9, 2026

12 remaining items

NotRexed commented on Nov 11, 2026

NotRexed commented on Nov 12, 2026

NikoMuffin commented on Nov 12, 2026

MidobiM-Code commented on Nov 12, 2026

mooki000 commented on Nov 12, 2026

NikoMuffin commented on Nov 12, 2026

Metadata

Metadata

Assignees

Labels

Projects

Milestone

Relationships

Development

Issue actions