compiled artefacts from win_x64 (sm_120)

README.md

@@ -113,9 +113,16 @@ Measured on one laptop: Intel Core Ultra 9 275HX / NVIDIA RTX 5060 Laptop, batch
 | lane | per batch | throughput | per word |
 |---|--:|--:|--:|
 | **TensorRT FP16, RTX 5060 Laptop** | **3.60 ms** | **17 790 words/s** | 0.056 ms |
-| **OpenVINO INT8, CPU** | **80.4 ms** | **796 words/s** | 1.26 ms |
+| **OpenVINO INT8, iGPU** (OV 2026.3) | **56.1 ms** | **1 140 words/s** | 0.88 ms |
+| **OpenVINO INT8, CPU** (OV 2026.3) | **75-83 ms** | **770-850 words/s** | 1.18-1.30 ms |
+| OpenVINO INT4, NPU (OV 2026.3) | 57.8 ms | 277 words/s | 3.61 ms |
 | ONNX Runtime FP32, CPU | 188 ms | 341 words/s | 2.93 ms |
 
+The NPU number is per batch 16, not 64 — it is a low-batch part, so its per-word cost is the
+highest of the three even though its per-batch latency looks similar to the CPU's. Running all
+three Intel devices at once yields ~87 % of the sum of their solo rates (they share one memory
+controller), so a host with no discrete GPU can still denoise ~2 000 words/s.
+
 The INT8 CPU lane is the intended default — 800 words/sec is enough to filter a repository's whole
 harvested vocabulary in seconds without touching a GPU, and it is 2.3× the unquantized ONNX path.
 The GPU lane exists for hosts that have spare VRAM anyway.
@@ -170,9 +177,9 @@ drops really was ballast.
 
 ## 7. What is in this repo
 
-- **OpenVINO INT8** — `code-daemon-denoise-v1-s_ov2026.2_{cpu,igpu_lnl}_int8_b64_s40.{xml,bin}` — the
+- **OpenVINO INT8** — `code-daemon-denoise-v1-s_ov2026.3_{cpu,igpu_lnl}_int8_b64_s40.{xml,bin}` — the
   default lane (CPU) and an Intel iGPU build.
-- **OpenVINO INT4, NPU** — `code-daemon-denoise-v1-s_ov2026.2_npu_int4_b16_s40.{xml,bin}` — weight-only
+- **OpenVINO INT4, NPU** — `code-daemon-denoise-v1-s_ov2026.3_npu_int4_b16_s40.{xml,bin}` — weight-only
   INT4 at batch 16 for Intel NPUs.
 - **TensorRT FP16** — `code-daemon-denoise-v1-s_{win_x64,linux_x64}_trt11.0_sm_120.engine`.
 - **TVM Vulkan** — `code-daemon-denoise-v1_{win_x64,linux_x64}_tvm0.25_vulkan.{dll,so}` — GPU fallback

manifest.json

@@ -10,10 +10,10 @@
   ],
   "runtime_tags": {
     "trt": "trt11.0",
-    "ov": "ov2026.2",
+    "ov": "ov2026.3",
     "tvm": "tvm0.25",
     "mlx_dir": "model_gpu_mlx0.22"
   },
-  "compiled_at": "2026-08-03T14:52:07Z",
+  "compiled_at": "2026-08-05T08:54:59Z",
   "compiled_by": "models/_compile"
 }