diff --git a/docs/Dankclaw/ComfyUI.md b/docs/Dankclaw/ComfyUI.md
index 5bbab2f..f52653f 100644
--- a/docs/Dankclaw/ComfyUI.md
+++ b/docs/Dankclaw/ComfyUI.md
@@ -16,7 +16,9 @@ ComfyUI is a node-based AI image generation engine. Workflows are built from nod
 
 ## Why CPU on Dankclaw
 
-Dankclaw is an AMD Strix Halo APU (gfx1151). Precompiled ROCm PyTorch wheels do not support this iGPU. This matches the existing knowledge base entry for ComfyUI on Dankclaw: the running instance uses `--cpu`. The Docker image for CPU execution is the `comfyui` CPU build; the GPU images (`comfyui-cuda`) must not be used.
+Dankclaw is an AMD Strix Halo APU (gfx1151). Precompiled PyTorch ROCm wheels do not ship kernels for this iGPU, and the official `rocm/comfyui` image is built for datacenter archs only (`PYTORCH_ROCM_ARCH=gfx942;gfx950`). This matches the knowledge base entry: the running instance uses `--cpu`.
+
+> **Update (2026-08-14):** the blanket claim "ROCm unavailable on Strix Halo" is now qualified. AMD publishes gfx1151-tagged PyTorch builds that work on this hardware, and sd.cpp/Lemonade ship gfx1151 kernels today. See [[ComfyUI ROCm and Vulkan]] and [[Lemonade and sd.cpp]]. The CPU config below remains correct until a gfx1151-enabled torch image is built.
 
 ## Command Line (from `cli_args.py`)
 
@@ -51,4 +53,6 @@ The web UI responds on the root path `/`. ComfyUI also exposes `/system_stats`.
 ## See Also
 
 - [[Dankclaw Stack]]
+- [[ComfyUI ROCm and Vulkan]] — live-verified GPU path research
+- [[Lemonade and sd.cpp]] — the working ggml-family alternative
 - Knowledge base: [[ComfyUI]] (tools), [[ComfyUI Troubleshooting]]
\ No newline at end of file
diff --git a/docs/Research/ComfyUI ROCm and Vulkan.md b/docs/Research/ComfyUI ROCm and Vulkan.md
new file mode 100644
index 0000000..45af34e
--- /dev/null
+++ b/docs/Research/ComfyUI ROCm and Vulkan.md	
@@ -0,0 +1,106 @@
+# ComfyUI on ROCm and Vulkan
+
+Research into whether ComfyUI can use GPU acceleration — not just CPU — on the AMD Strix Halo APU (gfx1151) in Dankclaw. Verified on live hardware 2026-08-14. This page corrects the persistent "ROCm unavailable on Strix Halo" claim.
+
+## TL;DR
+
+| Path | Works on gfx1151? | Evidence |
+|------|-------------------|----------|
+| pytorch.org ROCm wheels | **No** | wheels are built for `gfx906/gfx90a/gfx942/gfx1100` etc. — no gfx1151 kernels |
+| Official `rocm/comfyui` image | **No** | built with `PYTORCH_ROCM_ARCH=gfx942;gfx950` — datacenter archs only |
+| AMD gfx1151-tagged PyTorch build | **Yes** | `torch 2.9.1+rocm7.12.0rc1` → `avail: True`, `Radeon 8060S`, cap `(11,5)` |
+| Vulkan (torch.backends.vulkan) | **No** | compiled out of standard wheels; experimental/mobile even when present |
+| ComfyUI `--cpu` | **Yes (slow)** | current stack config |
+
+## The core insight
+
+PyTorch's ROCm support is not one thing. The wheels you install for poke at torch, the wheels AMD builds for specific APUs are different, and the kernels compiled in are fixed at build time via `PYTORCH_ROCM_ARCH`.
+
+Standard wheels (pytorch.org and the `rocm/comfyui` image) target **CDNA datacenter archs** (`gfx942`/`gfx950`) or the older common set. **gfx1151 (Strix Halo / Radeon 8060S)** is not included, so `torch.cuda.is_available()` returns `False` and everything falls back to CPU.
+
+AMD does publish gfx1151 builds — but you have to pick the right image.
+
+## Live verification (Dankclaw, 2026-08-14)
+
+### Fails: official rocm/comfyui image
+
+```bash
+podman pull docker.io/rocm/comfyui:comfyui-0.18.2.amd0_rocm7.2.0_ubuntu24.04
+podman run --rm --device=/dev/kfd --device=/dev/dri --entrypoint python <img> -c "
+import torch
+print(torch.__version__)          # 2.10.0a0+git…
+print(torch.cuda.is_available())  # False
+print(torch.cuda.get_arch_list()) # []"
+
+# Env in that image:
+#   PYTORCH_ROCM_ARCH=gfx942;gfx950
+#   ROCM_PATH=/opt/rocm-7.2.0
+```
+
+The image bundles ROCm 7.2 and compiles kernels only for `gfx942;gfx950`. No gfx1151. `HSA_OVERRIDE_GFX_VERSION` does not help because the kernels simply are not compiled in.
+
+The official ComfyUI ROCm image is 68 GB and unusable on Strix Halo. It was pulled on Dankclaw for this test; remove it to reclaim space.
+
+### Works: AMD gfx1151-tagged PyTorch build
+
+The `rocm/vllm` image exists in two flavors. The one tagged `gfx1151` works on Strix Halo:
+
+```bash
+podman run --rm \
+  --device=/dev/kfd --device=/dev/dri \
+  --entrypoint python \
+  docker.io/rocm/vllm:rocm7.12.0_gfx1151_ubuntu24.04_py3.12_pytorch_2.9.1_vllm_0.16.0 \
+  -c "
+import torch
+print(torch.__version__)        # 2.9.1+rocm7.12.0rc1
+print(torch.version.hip)        # 7.12.60610
+print(torch.cuda.is_available()) # True
+print(torch.cuda.get_device_name(0))  # Radeon 8060S Graphics
+print(torch.cuda.get_device_capability(0)) # (11, 5)  = gfx1151
+"
+
+# Env in that image:
+#   PYTORCH_ROCM_ARCH=gfx1151
+```
+
+Key facts:
+- `avail: True`, device `Radeon 8060S`, capability `(11, 5)` = gfx1151.
+- Device access requires passing `/dev/kfd` and `/dev/dri` to the container (and `--group-add video`/`realtime` or a user with those groups).
+- ROCm HIP runtime vars matter: this image has HIP 7.12; the older `rocm/vllm:rocm7.14.0_rdna_*` (48.6 GB) targets the newer "RDNA" set.
+
+## What this means for ComfyUI
+
+ComfyUI is a PyTorch app. If you give it a PyTorch wheel with gfx1151 kernels — the AMD `/opt/python` build used by the working vLLM image, or a custom build with `PYTORCH_ROCM_ARCH=gfx1151` — ComfyUI **will** use the GPU via ROCm. The blocker is not ComfyUI; it is shipping it in a container that contains a gfx1151-enabled torch.
+
+Three ways forward (future work, none applied yet):
+
+1. **Base a ComfyUI image on the AMD gfx1151 torch build.** Take `rocm/vllm:...gfx1151...pytorch_2.9.1...` (or AMD's PyTorch wheel index) and install ComfyUI on top. This is the direct path to GPU ComfyUI.
+2. **Build PyTorch for gfx1151** (`export PYTORCH_ROCM_ARCH=gfx1151` before build). Slow build, but a self-contained wheel afterwards.
+3. **Use a ggml-family runtime instead** — stable-diffusion.cpp / Lemonade's `sd.cpp` backend ship their own gfx1151 HIP kernels and work today (see [[Lemonade and sd.cpp]]).
+
+## Vulkan
+
+- PyTorch's Vulkan backend (`torch.backends.vulkan`) is **compiled out of standard Linux wheels**. The installed `torch 2.11/2.13` on Dankclaw reports `hasattr(torch.backends, 'vulkan')` → `False`.
+- Even when built in, PyTorch Vulkan is a mobile-oriented experimental backend, not a desktop inference path, and ComfyUI exposes no `--vulkan` flag.
+- The system has a working Vulkan driver (RADV, `libvulkan_radeon.so` + `radeon_icd.json`), so Vulkan apps work — but ComfyUI/PyTorch cannot consume it here.
+- Conclusion: **Vulkan is a dead end for ComfyUI.** If GPU is wanted, ROCm is the realistic path; a ggml-family engine (sd.cpp/Lemonade) is the pragmatic one.
+
+## Consequence for this repo
+
+The current `dankclaw/compose.yaml` runs ComfyUI with `--cpu`. That remains correct **until a gfx1151-enabled torch image is built**. When GPU priority arrives, swap the image for one based on the AMD Strix-savvy build and drop `--cpu`.
+
+## Cleanup note
+
+These images were pulled on Dankclaw during this research:
+
+```bash
+podman rmi docker.io/rocm/comfyui:comfyui-0.18.2.amd0_rocm7.2.0_ubuntu24.04   # 68 GB, unusable on gfx1151
+podman rmi docker.io/rocm/vllm:rocm7.12.0_gfx1151_...                         # 32.5 GB, keep if pursuing ComfyUI ROCm
+```
+
+## See Also
+
+- [[ComfyUI]] — the service page (CPU config)
+- [[Lemonade and sd.cpp]] — the ggml-family alternative with working gfx1151 kernels
+- [[Observability]], [[Dankclaw Stack]]
+- KB: [[ComfyUI]] (tools), [[ComfyUI Troubleshooting]] (the "ROCm unavailable" claims here are now qualified)
\ No newline at end of file
diff --git a/docs/Research/Lemonade and sd.cpp.md b/docs/Research/Lemonade and sd.cpp.md
new file mode 100644
index 0000000..a46d8ba
--- /dev/null
+++ b/docs/Research/Lemonade and sd.cpp.md	
@@ -0,0 +1,49 @@
+# Lemonade and sd.cpp
+
+Why Lemonade Server can do image generation on Strix Halo when ComfyUI cannot. Research from the Lemonade source repo (`lemonade-sdk/lemonade`, cloned 2026-08-14) and live verification.
+
+## The short answer
+
+Lemonade does not run ComfyUI. Its image-generation backend is a wrapped **stable-diffusion.cpp** (`leejet/stable-diffusion.cpp`) subprocess. sd.cpp is a ggml-family native C++ runtime that ships its own HIP kernels — and those kernels **include gfx1151**. ComfyUI runs on PyTorch, whose standard wheels omit gfx1151. Same hardware, different runtime, different result.
+
+## From the Lemonade source
+
+`src/cpp/server/backends/sdcpp/sdcpp_server.cpp`:
+
+- Image gen is `SDServer`, a `WrappedServer` around a prebuilt sd.cpp binary.
+- Downloads `sd-<version>-bin-Linux-Ubuntu-24.04-x86_64-<backend>.zip` from `leejet/stable-diffusion.cpp` releases.
+- Backends: `cpu, cuda, metal, rocm, vulkan`.
+- ROCm asset is `...-rocm-<version>.zip`; Vulkan is `...-vulkan.zip`.
+
+Backend support matrix (`docs/dev/backends-reference.md`):
+
+| Recipe | Backend | Device families |
+|--------|---------|-----------------|
+| `sd-cpp` | rocm | amd_gpu (gfx103X, gfx110X, **gfx1150, gfx1151, gfx1152**, gfx120X) |
+| `sd-cpp` | vulkan | amd_gpu; cpu; nvidia_gpu |
+| `thenoise` | rocm | amd_gpu (gfx1150, gfx1151, gfx1152) — experimental |
+
+**gfx1151 is explicitly supported** by sd.cpp's ROCm build. There is also a dedicated experimental `thenoise` ROCm backend targeting exactly the Strix Halo archs.
+
+## Engine comparison
+
+| | ComfyUI | Lemonade (sd.cpp) |
+|---|---|---|
+| Runtime | PyTorch | native C++ (ggml family) |
+| AMD path | needs PyTorch ROCm wheel with gfx1151 kernels — not in standard wheels | ships own HIP kernels, gfx1151 included |
+| On Strix Halo | CPU only (today) | ROCm GPU, or Vulkan fallback |
+| Interface | node graph web UI | server, OpenAI-style API, GUI/CLI |
+
+This mirrors llama.cpp on the same box: ggml-family runtimes ship their own gfx1151 HIP kernels and run on ROCm; PyTorch does not.
+
+## Relevance
+
+- If GPU image generation on Dankclaw matters, **Lemonade (or bare sd.cpp) is the working path today**.
+- ComfyUI's node graph is the cost of switching: Lemonade is not ComfyUI.
+- See [[ComfyUI ROCm and Vulkan]] for the PyTorch-side analysis showing a gfx1151-enabled torch would also make ComfyUI work.
+
+## See Also
+
+- [[ComfyUI ROCm and Vulkan]] — the PyTorch side of this question
+- [[ComfyUI]] — current CPU config
+- [[Dankclaw Stack]]
\ No newline at end of file
diff --git a/docs/index.md b/docs/index.md
index 2bd8ac7..091eb65 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -55,6 +55,8 @@ This repository tracks the container stacks for both hosts. Git is the single so
 ## Research
 
 - [[Service Research Brief]] — consolidated subagent research behind these pages
+- [[ComfyUI ROCm and Vulkan]] — live-verified GPU path for ComfyUI on Strix Halo
+- [[Lemonade and sd.cpp]] — the ggml-family image-generation alternative
 - [[Backblaze Integration]] — off-site backup plan (future)
 
 ## See Also