moe: sparse top-k decode — compute only routed experts (1.8x, beats llama TP=2)

Dense MoE replicated x across all 16 local experts and ran the full
batched GEMM, reading every expert's weights per token; the weighted
sum then discarded 12 of 16 results. Decode is memory-bound, so this
was ~8x wasted expert bytes — the entire decode gap vs llama.cpp.

New fused expert-indexed GEMVs (csrc/moe/moe_sparse.cu) read
topk_ids on-device (no host sync) and early-return block-uniformly
for experts other ranks own. FP8 runs W8A16 (activations stay BF16 —
tensor cores are irrelevant at M=1, and activation quantization error
disappears); MXFP4 runs W4A16. Per-expert bias + scale fused into the
GEMV epilogue; slot-indexed weighted sum skips (never multiplies)
unwritten non-local slots. Dense path retained for num_tokens > 8
(prefill) and via XSERV_DENSE_MOE=1 for A/B.

dash5 (RTX 5090), gpt-oss-20b FP8, TP=2: decode TPOT 13.9 -> 7.6 ms.
Warm-server vs llama.cpp MXFP4 TP=2: TPOT 7.19-7.32 vs 7.54-8.42 ms —
first config where xserv wins decode outright. GSM8K-100: 96% (dense
FP8: 91%). llama TP=1 (2.9 ms) remains ahead: next levers are decode
CUDA graphs, non-expert quantization, sparse prefill (docs/20).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

crates/xserv-kernels/build.rs

@@ -31,6 +31,7 @@ fn main() {
         .file("../../csrc/attention/paged_attention.cu")
         .file("../../csrc/attention/reshape_and_cache.cu")
         .file("../../csrc/moe/moe_kernels.cu")
+        .file("../../csrc/moe/moe_sparse.cu")
         .file("../../csrc/quantization/dequant_fp8.cu")
         .file("../../csrc/quantization/quantize_fp8.cu")
         .file("../../csrc/quantization/mxfp4_gemm.cu")