Bump FlashInfer to 0.6.17 and remove Kimi K3 workarounds (#33997)

docs/cookbook/autoregressive/Moonshotai/Kimi-K3.mdx

@@ -127,16 +127,7 @@ Capacity levers, all in the Playground. Each trades precision or cache behavior
 
 Speculation: DSPARK holds block size + 1 (= 8) intermediate states per request — the calculator folds this in — and an unset `--max-running-requests` resets to 48 under spec (the command panel reminds you; set it explicitly to raise).
 
-**MoE runner.** Leave `--moe-runner-backend` unset on Blackwell and it resolves to FlashInfer MXFP4 (W4A8, prebuilt trtllm-gen SiTU kernels) when the cubin pool is installed, Marlin (W4A16) otherwise; H100/H200 pin Marlin. The B200 Balanced and High-Throughput cells pin `flashinfer_mxfp4` explicitly because that is the shape they were brought up on — on an install without the pool, drop the flag to fall back to Marlin. The published Docker images already provision the **SiTU cubin pool**; to install it independently, run the same flow as the Dockerfile:
-
-```bash
-wget https://github.com/sgl-project/whl/releases/download/trtllm_gen_moe_cubin_20260617/trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip
-sudo mkdir -p /opt/trtllm_gen_moe_cubin_pool
-sudo unzip -q trtllm_gen_moe_cubin_pool_20260617_v0613rc1.zip -d /opt/trtllm_gen_moe_cubin_pool
-export SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL=/opt/trtllm_gen_moe_cubin_pool/trtllm_gen_moe_cubin_pool_20260617_v0613rc1
-```
-
-Remaining kernel sources JIT once from the public `flashinfer` wheel (a few minutes, cached).
+**MoE runner.** Leave `--moe-runner-backend` unset on Blackwell: FlashInfer MXFP4 (W4A8, official trtllm-gen SiTU kernels) is selected with the pinned FlashInfer 0.6.17 dependency; H100/H200 pin Marlin. The B200 Balanced and High-Throughput cells pin `flashinfer_mxfp4` explicitly because that is the shape they were brought up on. The published Docker images install the matching official `flashinfer-python`, `flashinfer-cubin`, and `flashinfer-jit-cache` packages.
 
 **Attention backend.** Leave all three attention knobs unset on Blackwell: K3 resolves prefill, decode, and — under DSPARK — verification as a set (`trtllm_mla` across the board; `cutedsl_mla` takes decode and verification under DCP). On the non-DCP recipes, setting any one of the three cancels the auto-resolution for the others. The B200 Balanced and High-Throughput cells pin `--decode-attention-backend cutedsl_mla`, which is what auto-resolution picks for those DCP recipes anyway — it is written out because it is the shape they were brought up on, not because it changes the resolution. H100/H200 pin `flashmla` for decode.
 
@@ -395,7 +386,7 @@ Both presets are one click away in the [Playground above](#playground): pick a *
 
 Decisions the preset already makes:
 
-- **MegaMoE on `deep_gemm`** — the fastest a2a backend; needs the SiTU cubin pool ([§2](#2-configuration-tips)).
+- **MegaMoE + `deep_gemm`** — the fused DeepGEMM all-to-all/MoE path used by these large-scale DP/EP throughput presets, with K3's SiTU activation.
 - **SP-MoE and shared-expert overlap** engage automatically under EP a2a; the K3 all-reduce fusion does not.
 - **Spec Decode follows the Deploy knob.** Acceptance thins at large batch; spec × EP × DP-attention is validated only at 8-GPU EP8 × DP2 (full GSM8K) — experimental at these scales.
 

docs/src/snippets/_playground.jsx

@@ -629,7 +629,7 @@ export const Playground = ({ config }) => {
             "--moe-a2a-backend", "--moe-runner-backend",
           ]);
           // Backend options may carry their own env (e.g. the FlashInfer MXFP4
-          // cubin-pool path): strip every backend option's env keys, then
+          // backend-specific path): strip every backend option's env keys, then
           // re-add the selected option's.
           const backendEnvKeys = [];
           for (const o of (fc.backend?.options || [])) {

docs/src/snippets/configs/moonshotai/kimi-k3.jsx

@@ -458,10 +458,8 @@ export const config = {
           // Blackwell-only kernel-fusion path; selecting it reveals the Quantization sub-select.
           { id: "megamoe",          label: "MegaMoE",           flags: ["--moe-a2a-backend megamoe"],
             requiresHw: ["b200", "b300", "gb200", "gb300"] },
-          // Blackwell-only: runs the prebuilt trtllm-gen SiTU cubins; needs the
-          // downloadable SiTU cubin pool unpacked and pointed to by the env var.
+          // Blackwell-only: runs FlashInfer's official trtllm-gen SiTU kernels.
           { id: "flashinfer_mxfp4", label: "FlashInfer (MXFP4)", flags: ["--moe-runner-backend flashinfer_mxfp4"],
-            env: ["SGLANG_TRTLLM_GEN_MOE_CUBIN_POOL=/path/to/trtllm_gen_moe_cubin_pool"],
             requiresHw: ["b200", "b300", "gb200", "gb300"] },
           { id: "marlin",           label: "Marlin (W4A16)",    flags: ["--moe-runner-backend marlin"] },
         ],
@@ -931,10 +929,8 @@ export const config = {
         "--pp-size 2",
         "--dcp-size 8",
         "--ep-size 8",
-        // Both pinned to the brought-up shape rather than left to the auto
-        // resolution the rest of Blackwell uses. The MXFP4 runner needs the SiTU
-        // cubin pool the published image ships; drop it to get the Marlin
-        // fallback on an install without one.
+        // Both backends are pinned to the validated recipe even though automatic
+        // resolution selects them on Blackwell.
         "--moe-runner-backend flashinfer_mxfp4",
         "--decode-attention-backend cutedsl_mla",
         "--mem-fraction-static 0.85",