sycl: STRATA_PF_XMX=2 is the default on Arc Alchemist (opt-out STRATA_PF_XMX=0); Battlemage unchanged

A750 prompt prefill +30..39% at 4K (mode 2: 79.3 -> 110.4 tok/s, 5 of 5 rounds), +32..35% at 1-2K tokens; decode and cache slots unchanged.
Teacher-forced KL against the default 0.0002 (floor 0.0001), first-token KL 0.0017 (floor 0.0005), tasks inside the default's run-to-run spread.
Docs: INTEL.md has the numbers.

docs/INTEL.md

@@ -759,6 +759,17 @@ Measured on an Arc A750 (8 GB, `i915`, PCIe 4.0) with the Flash-Next IQ3_XXS in
   8192 --prefill-repeat 5` for the time of both kernels. Not taken: the same kernel with its state in SLM (no register
   spills, the same bits, ~4x slower: 13.7 vs 3.1-3.3 ms at T=1000).
 
+### Prompt speed on Alchemist: the grouped XMX expert launch is the default (0.1.48, `STRATA_PF_XMX`)
+
+The A750's prompt path decodes and multiplies a group of experts in one fused launch on the XMX units (`STRATA_PF_XMX=2`, `sycl/src/prefill/xmx_moe.dp.cpp`)
+instead of dequantizing each expert to FP16 and calling oneMKL. It is the default on Alchemist cards; `STRATA_PF_XMX=0` turns it off and `=1` is the
+variant that dequantizes first and runs one grouped XMX GEMM per product. Battlemage keeps the old path (the same kernels are 7-11% slower there).
+Measured on the A750 (IQ3_XXS, default flags, interleaved rounds against the run before): a 1K-token prompt 35.1 -> 46.4 tok/s, 2K 61.7 -> 83.1, 4K 79.3 -> 110.4
+(+39%, 5 of 5 rounds), 6K 66.3 -> 114.8; decode unchanged (18.7 tok/s), the same number of cache slots. It rounds differently from the FP16 path:
+teacher-forced KL against the default over 6,006 forced tokens 0.0002 (the default against itself: 0.0001; with the default PCIe share the A750's own run-to-run noise
+is 0.012), first-token KL 0.0017 (floor 0.0005), top-1 agreement 99.7%. HumanEval (every 2nd, 82) 73, 69 and 74 for the mode and two default runs, reasoning 21 / 23 / 22 of
+30, a 35-question Chinese/Japanese/Korean set 35/35 in all, needles at 2K-6K 9/9, an 8K-context tool-call agent 6/6.
+
 ## Not done
 
 - Images, on both Intel engines. Strata's vision path encodes with `strata-vision` into embeddings the CUDA

sycl/src/prefill/xmx_moe.dp.cpp

@@ -150,7 +150,11 @@ int mode() {
     // sub-group 16 (Xe2, Battlemage).
     static const int v = [] {
         const char* e = std::getenv("STRATA_PF_XMX");
-        const int want = e ? std::atoi(e) : 0;
+        // 0.1.48: mode 2 (the fused grouped XMX launch) is the default on Arc Alchemist (A-series): A750 prompts +30..39% (4K), +32..35% (1-2K tokens),
+        // first-token KL 0.002 against a 0.0005 floor, HumanEval / reasoning / agent / CJK / needles inside the run-to-run spread; Battlemage is
+        // 7-11% slower with it, so it stays off there.  STRATA_PF_XMX=0 turns it off, 1 is the dequantize + grouped XMX variant.
+        const bool alchemist = strata::intel_gpu_gen(strata::q_of(nullptr)->get_device()) == strata::IntelGpuGen::Alchemist;
+        const int want = e ? std::atoi(e) : alchemist ? 2 : 0;
         if (want == 0) return 0;
 #ifdef STRATA_NO_DG2_XMX
         std::fprintf(stderr, "STRATA_PF_XMX ignored: this build is AOT for a device without the sub-group 8 XMX kernels\n");