Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
C++
RX 6000 package: docs, release-notes draft, Windows build comment, tuning-table test rows
Niko1221ac71d2b32e95
diff --git a/docs/AMD_HIP.md b/docs/AMD_HIP.mdindex 5841adb65..97313ab7c 100644--- a/docs/AMD_HIP.md+++ b/docs/AMD_HIP.md@@ -73,9 +73,11 @@ in Settings (or run `strata update`) and the matching package downloads.| --- | --- | --- | --- | --- || `rx9070` | Windows | gfx1200, gfx1201 | 10.2.0a20260930 (TheRock) | `Strata-0.2.1-windows-hip.exe` || `rdna3` | Windows | gfx1100, gfx1101, gfx1102, gfx1151 | 10.2.0a20260930 | `Strata-0.2.1-windows-hip-rdna3.exe` |+| `rdna2` (0.2.2) | Windows | gfx1030, gfx1031 | 10.2.0a20260930 | no installer of its own: any app installer, then Update engine || `rx9070` | Linux | gfx1200, gfx1201 | 7.10.0a20251120 (`gfx120X-all` wheels) | `Strata-0.2.1-linux-amd-rx9070.AppImage` || `rx7900` | Linux | gfx1100, gfx1101, gfx1102 | 7.10.0a20251120 (`gfx110X-dgpu` wheels) | `Strata-0.2.1-linux-amd-rx7900.AppImage` || `halo` | Linux | gfx1151 | 7.14.1 (TheRock) | `Strata-0.2.1-linux-amd-strixhalo.AppImage` |+| `rx6000` (0.2.2) | Linux | gfx1030, gfx1031 | 7.13.0a20260515 (`gfx103X-all` wheels) | no installer of its own: any app installer, then Update engine |`rdna3` and `halo` were added to the release after the first upload (feed sequence 101). Why the split: on Windows one ROCm 10.2 libraryset holds the kernels of every listed architecture, so RX 7000 and Strix Halo share a package. The Linux 7.10 wheels hold one chip family@@ -84,8 +86,18 @@ per library set, and 7.10 crashes in the HSA runtime on a Strix Halo with a newEach package carries the engine, the CPU picture reader and the GPU picture reader (`strata-vision-hip`) for the same architectures, andthe rocBLAS and hipBLASLt kernel data for them only. A package whose GPU code does not cover the card is refused with one line thatnames the card family ("This build of Strata is for AMD RX 9000 cards. This computer has an AMD RX 7000 card ...").-**The RX 6000 series (gfx1030, RX 6800 / 6900) is in no package of this release**, on Windows or Linux: the app says so plainly instead of naming a-build to ask for. An engine built for it ([RDNA2](#rdna2-gfx1030), `STRATA_HIP_ARCHS=gfx1030`) still runs it.+**The RX 6000 series (gfx1030 / gfx1031, RX 6800 / 6900 / 6700 XT) has its own package from 0.2.2** (below). The 0.2.1 packages have no code for it,+and the app said so plainly instead of naming a build to ask for.++**RX 6000 package (0.2.2): built, not tested on a real RX 6000 card.** `rx6000` on Linux (ROCm 7.13.0a20260515, the `gfx103X-all` wheels+setup already uses for these cards) and `rdna2` on Windows (ROCm 10.2.0a20260930), each with the code for gfx1030 and gfx1031 only. It is a niche,+opt-in build: the feed offers it to a PC whose card reports gfx1030 or gfx1031 and to no other (a PC with an RX 7000 / 9000 card beside an RX 6000+one keeps the package it always had), and the RX 7000 / 9000 / Strix Halo packages are unchanged. It runs the plain+kernels: RDNA2 has no matrix cores, so none of the WMMA paths, the gfx12 prompt defaults or the 780M prompt set apply (they are chosen by the card's+architecture, `src/core/arch_defaults.cpp`), and hipBLASLt ships no gfx1030 kernels, so there is no tuning table and the dense matrix products use plain+hipBLAS. Linux build: `VARIANT=rx6000 packaging/build_hip_linux.sh`; Windows: `STRATA_HIP_ARCHS="gfx1030;gfx1031" tools\hip\build_windows.bat`, then+`package_engine.py --variant rdna2`. The [RDNA2 section](#rdna2-gfx1030) has what community owners measured with engines built from source. Please+report a run on an RX 6000 card with the end of `strata-<model>.log`.**Automatic card (0.2.1):** with no card named in the model's settings, the engine start asks the package's `strata-device --list-devices`and skips cards the package has no code for and integrated chips (a mini PC's Radeon 760M / 890M is HIP's device 0) while a card of its own@@ -107,7 +119,7 @@ What was run on real hardware:- **Tuning tables (0.2.1):** each AMD package carries the hipBLASLt tuning tables of `tools/hip` that match the hipBLASLt it bundles, as`tuning/<arch>-hipblaslt-<version>.txt` (listed and hashed in the package manifest), and records that version as `hipblaslt_version` in itsBUILD.json. Windows `rx9070` has the gfx1201 table, Windows `rdna3` the gfx1100 and gfx1151 ones (all hipBLASLt 1.5.0), Linux `rx7900` the- gfx1100 one (1.2.0), Linux `halo` the gfx1151 one (1.4.1). Linux `rx9070` has none: its wheels carry hipBLASLt 1.2.0 and no table in `tools/hip`+ gfx1100 one (1.2.0), Linux `halo` the gfx1151 one (1.4.1). The RX 6000 packages (`rx6000`, `rdna2`, 0.2.2) have none, because hipBLASLt has no gfx1030 / gfx1031 kernels. Linux `rx9070` has none: its wheels carry hipBLASLt 1.2.0 and no table in `tools/hip`is for that version (a table calibrated for it on an R9700 read prompts about 1% faster and decode 2% slower, so none is shipped).The engine start picks the table for the card and uses it only when the package's hipBLASLt version is the table's; otherwise it logswhy and the dense matrix products stay on plain hipBLAS. **A table changes which GEMM algorithm runs the prompt, so the rounding differs: thediff --git a/docs/release-notes-0.2.2.md b/docs/release-notes-0.2.2.mdnew file mode 100644index 000000000..5d6794b29--- /dev/null+++ b/docs/release-notes-0.2.2.md@@ -0,0 +1,5 @@+**0.2.2 (draft).** Only the items of the `f-022-rx6000` and `f-022-numa` branches are written here; the coordinator adds the rest.++## What is new++- **An engine package for the RX 6000 series (built, not tested on a real RX 6000 card).** Linux has `rx6000` and Windows has `rdna2`, each with GPU code for gfx1030 and gfx1031 (RX 6800, 6900 and 6950 series, RX 6700 XT). They are offered to a PC whose card reports one of those two and to no other: the RX 7000, RX 9000 and Strix Halo packages are unchanged, and a PC with an RX 7000 or 9000 card next to an RX 6000 card keeps the package it had. The engine runs plain kernels on these cards (RDNA2 has no matrix cores, so none of the RX 7000 / RX 9000 or Radeon 780M prompt paths apply) and plain hipBLAS (hipBLASLt ships no kernels for gfx1030, so there is no tuning table). Setup and the app say once that this package was built and not tested on a real RX 6000 card. What was checked: the Linux package builds, every program in it has code for exactly gfx1030 and gfx1031, the rocBLAS kernel files for both are in it, every library it imports is in the package or belongs to the system, and the whole engine (all targets, tests included) compiles for gfx1030 and gfx1031. What was not: a model on an RX 6000 card, on either system, and the Windows package itself (the Windows build scripts take the new architecture list, but no Windows `rdna2` package has been built). Earlier reports on these cards came from people who built the engine from source ([docs/AMD_HIP.md](AMD_HIP.md#rdna2-gfx1030)). If something goes wrong, open an issue with the end of `strata-<model>.log`.diff --git a/packaging/ENGINE-COMPONENT.md b/packaging/ENGINE-COMPONENT.mdindex 100811927..db7c12215 100644--- a/packaging/ENGINE-COMPONENT.md+++ b/packaging/ENGINE-COMPONENT.md@@ -101,6 +101,7 @@ the manifest file (which holds every file's hash), used as the package's digest| `windows-x86_64` hip | the owner PC, `tools/hip/build_windows.bat` (ROCm 10.2 nightly wheels) then `tools/hip/package_windows.py` | `strata.exe`, `strata-device.exe`, the ROCm DLLs and the rocBLAS/hipBLASLt kernels of the zip (`--tree`) | no Windows AMD test machine yet; with `STRATA_NO_VISION` unset it also builds the two picture readers (`strata-vision.exe` on the CPU, `strata-vision-hip.exe` on the GPU) || `linux-x86_64` cuda | a Linux box with the CUDA toolkit | `strata`, `libstrata_kq.so`, `libcudart/libcublas/libcublasLt` (13.x) | the GPU driver's `libcuda` stays the PC's || `linux-x86_64` hip | a box with ROCm (the Python wheels are enough) | `strata`, the HIP runtime, hipBLAS/hipBLASLt/rocBLAS, comgr + libLLVM, `rocblas/library` and `hipblaslt/library` for the listed gfx archs; `packaging/build_hip_linux.sh` builds it with both picture readers |+| `linux-x86_64` hip, variant `rx6000` (RX 6000, gfx1030 / gfx1031) | `VARIANT=rx6000 packaging/build_hip_linux.sh` in the jammy chroot, ROCm 7.13.0a20260515 `gfx103X-all` wheels | the same set as the other HIP packages, with `rocblas/library` for gfx1030 / gfx1031; no `hipblaslt/library` kernels exist for them and no `tuning/` table | built, not tested on a real RX 6000 card (the manifest notes say so). `windows-x86_64` hip, variant `rdna2`: the same archs from `tools/hip/build_windows.bat` with `STRATA_HIP_ARCHS=gfx1030;gfx1031`, packaged with `--variant rdna2`. A PC with an RX 7000 / 9000 card beside an RX 6000 one never takes it (`engine_component.hip_preference`) || `linux-x86_64` hip, gfx1103 (Radeon 780M / Phoenix iGPU) | `packaging/build_hip_gfx1103.sh` on the mini PC (TheRock gfx110X-all SDK) | the same set as the gfx1200/1201 package, with `rocblas/library` and `hipblaslt/library` for gfx1103 (600 MiB) | experimental, unvalidated: the iGPU resets now and then (MES REMOVE_QUEUE); never kill the engine mid-request; run configs set `STRATA_NO_ARENA_THP=1`. `engine_component.resolve` picks a HIP package only when its `archs` list a gfx target the Linux kfd topology reports || `linux-x86_64` sycl | a box with oneAPI DPC++ | `strata` (JIT from SPIR-V), libsycl, the Unified Runtime adapters, MKL SYCL libraries | the Level Zero loader and compute runtime are the PC's driver || `macos-arm64` metal | an Apple Silicon Mac | `strata`, `metal/kernels/*.metal`, `metal/engine-msl/*.metal`, `metal/ggml-common.h`, `metal/data/*.bin` | the Metal engine compiles its kernel sources when it starts; since S1 it finds them beside the executable (`metal/src/resource_path.h`: environment variable, then `<exe dir>/metal/`, then the compiled-in source path) |diff --git a/packaging/README.md b/packaging/README.mdindex e85839895..d316ef445 100644--- a/packaging/README.md+++ b/packaging/README.md@@ -21,7 +21,7 @@ A backend name alone is not a target. Each row is an OS x CPU ABI x backend tupl| `windows-x86_64-cuda` | NSIS `.exe` (Tauri) | `cuda13` and `cuda12` packages | **built and tested** on a Windows PC with an NVIDIA card: silent install with a PATH of System32 only, a real model, an agent chat. || `windows-x86_64-hip` | NSIS `.exe` (Tauri) | `tools/hip/build_windows.bat` | **built; never run on a real AMD card.** || `linux-x86_64-cuda` | AppImage (Tauri) | `setup.py`; built in a glibc 2.35 chroot | **cuda13 built and tested** (RTX 3060 and RTX A4000: app in the browser, a real model, a picture). cuda12 built. |-| `linux-x86_64-hip` | AppImage (Tauri) | `packaging/build_hip_linux.sh` | **RX 9000 (gfx1200 / gfx1201) built and tested** on a Radeon AI PRO R9700. **RX 7000 (gfx1100 / 1101 / 1102) built; not run on an RX 7000 card.** |+| `linux-x86_64-hip` | AppImage (Tauri) | `packaging/build_hip_linux.sh` | **RX 9000 (gfx1200 / gfx1201) built and tested** on a Radeon AI PRO R9700. **RX 7000 (gfx1100 / 1101 / 1102) built; not run on an RX 7000 card.** **RX 6000 (gfx1030 / 1031, 0.2.2, engine package `rx6000` only) built; not tested on a real RX 6000 card.** || `linux-x86_64-sycl` | | `sycl/setup_intel.py` | Documented as experimental ([INTEL.md](../docs/INTEL.md)). || `windows-x86_64-sycl` | none | none | **gated.** No native Windows SYCL path exists upstream; it needs a separate feasibility gate, not a recipe. || `macos-x86_64-*` | none | none | **unsupported.** The macOS engine is Apple-Silicon only (`metal/CMakeLists.txt` requires Apple Silicon for the Metal runtime). |diff --git a/tests/terminal/test_hipblaslt_tuning.py b/tests/terminal/test_hipblaslt_tuning.pyindex 787b6d680..999ef2b2f 100644--- a/tests/terminal/test_hipblaslt_tuning.py+++ b/tests/terminal/test_hipblaslt_tuning.py@@ -30,6 +30,9 @@ PACKAGES = {("linux-x86_64", "rx9070"): (("gfx1200", "gfx1201"), 100200, set()), # no table for 1.2.0 on RDNA4 (docs/AMD_HIP.md)("linux-x86_64", "rx7900"): (("gfx1100", "gfx1101", "gfx1102"), 100200, {"gfx1100"}),("linux-x86_64", "halo"): (("gfx1151",), 100401, {"gfx1151"}),+ # 0.2.2 RX 6000 packages: hipBLASLt ships no gfx1030 / gfx1031 kernels at any version, so no table, plain hipBLAS+ ("windows-x86_64", "rdna2"): (("gfx1030", "gfx1031"), 100500, set()),+ ("linux-x86_64", "rx6000"): (("gfx1030", "gfx1031"), 100400, set()),}diff --git a/tools/hip/build_windows.bat b/tools/hip/build_windows.batindex 8f74e5b5c..1f7b3e10a 100644--- a/tools/hip/build_windows.bat+++ b/tools/hip/build_windows.bat@@ -13,6 +13,8 @@ remrem Settings (environment variables, all optional):rem STRATA_HIP_ARCHS gfx1100;gfx1101;gfx1102;gfx1200;gfx1201;gfx1030;gfx1031;gfx1151 (the cards setup supports, + gfx1102; gfx1151 = Strix Halo;rem gfx1031 = RX 6700 XT / 6800M, #915 #1078; drop it from the list to build the zip without it)+rem The release packages are built one per card family: rx9070 = gfx1200;gfx1201, rdna3 = gfx1100;gfx1101;gfx1102;gfx1151,+rem rdna2 (0.2.2, RX 6000: built, not tested on a real card) = gfx1030;gfx1031; each then goes through package_engine.py --variant <name>rem STRATA_ROCM_VERSION 10.2.0a20260930 STRATA_ROCM_INDEX https://nightly.repo.amd.com/rocm/whl-next/rem ROCM_VENV <repo>\.rocm-win BUILD_DIR <repo>\build-hip-win DIST_DIR <repo>\distrem STRATA_GGML_DIR a llama.cpp checkout at the pinned commit (default: CMake fetches it)