SemiAnalysisAI · cquil11 · Aug 12, 2026 · Aug 12, 2026 · Aug 12, 2026 · Aug 12, 2026
diff --git a/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh b/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh
@@ -0,0 +1,128 @@
+#!/usr/bin/env bash
+set -euo pipefail
+set -x
+
+# AgentX trace replay for Qwen3.5-397B-A17B MXFP4 on MI355X with SGLang
+# native EAGLE MTP. Throughput uses the committed golden synthetic
+# acceptance length; evaluation retains real target-model verification.
+
+source "$(dirname "$0")/../../benchmark_lib.sh"
+
+export EVAL_FRAMEWORK="lm-eval"
+
+check_env_vars \
+    MODEL TP CONC EP_SIZE RESULT_DIR DURATION
+
+SCHEDULER_RECV_INTERVAL=${SCHEDULER_RECV_INTERVAL:-30}
+
+if [[ -n "${SLURM_JOB_ID:-}" ]]; then
+    echo "JOB $SLURM_JOB_ID running on ${SLURMD_NODENAME:-unknown}"
+fi
+
+if [[ -n "${MODEL_PATH:-}" ]]; then
+    if [[ ! -d "$MODEL_PATH" || -z "$(ls -A "$MODEL_PATH" 2>/dev/null)" ]]; then
+        hf download "$MODEL" --local-dir "$MODEL_PATH"
+    fi
+else
+    hf download "$MODEL"
+    export MODEL_PATH="$MODEL"
+fi
+
+rocm-smi || true
+amd-smi || true
+
+export WEKA_LOADER_OVERRIDE=semianalysis_cc_traces_weka_062126_256k
+resolve_trace_source
+install_agentic_deps
+
+export AIPERF_SERVER_METRICS_URLS="http://localhost:${PORT}/metrics"
+export AIPERF_REQUIRED_SERVER_METRIC_PREFIX="sglang:"
+
+SERVER_LOG="$RESULT_DIR/server.log"
+mkdir -p "$RESULT_DIR"
+
+SERVER_PID=""
+cleanup_agentic_services() {
+    local exit_code=$?
+    trap - EXIT INT TERM
+    set +e
+    stop_background_process_tree "$SERVER_PID" "SGLang server" 60
+    exit "$exit_code"
+}
+trap cleanup_agentic_services EXIT
+trap 'exit 130' INT
+trap 'exit 143' TERM
+
+PARALLEL_ARGS=(
+    --tp "$TP"
+    --dp 1
+    --ep-size "$EP_SIZE"
+)
+
+TOKENIZER_ARGS=()
+if [ "$TP" -ge 4 ]; then
+    TOKENIZER_ARGS=(--tokenizer-worker-num 6)
+fi
+
+MAX_RUNNING_REQUESTS=$((2 * CONC))
+CUDA_GRAPH_MAX_BS="$CONC"
+[ "$CUDA_GRAPH_MAX_BS" -gt 64 ] && CUDA_GRAPH_MAX_BS=64
+
+export PYTHONNOUSERSITE=1
+export SGLANG_USE_AITER=1
+export SGLANG_USE_AITER_UNIFIED_ATTN=1
+export AITER_FLYDSL_FORCE=1
+export SGLANG_MAMBA_SSM_DTYPE=bfloat16
+export SGLANG_TIMEOUT_KEEP_ALIVE=1800
+
+if [ "${EVAL_ONLY:-false}" != "true" ]; then
+    export SGLANG_SIMULATE_ACC_LEN=3.39
+    export SGLANG_SIMULATE_ACC_METHOD=match-expected
+    export SGLANG_SIMULATE_ACC_TOKEN_MODE=real-draft-token
+fi
+
+SGLANG_CMD=(
+    python3 -m sglang.launch_server
+    --model-path "$MODEL_PATH"
+    --served-model-name "$MODEL"
+    --host 0.0.0.0
+    --port "$PORT"
+    --trust-remote-code
+    "${PARALLEL_ARGS[@]}"
+    --attention-backend aiter
+    --mem-fraction-static 0.80
+    --model-loader-extra-config '{"enable_multithread_load": true}'
+    --watchdog-timeout 1200
+    --page-size 16
+    --cuda-graph-max-bs "$CUDA_GRAPH_MAX_BS"
+    --max-running-requests "$MAX_RUNNING_REQUESTS"
+    --max-prefill-tokens 32768
+    --chunked-prefill-size 32768
+    --scheduler-recv-interval "$SCHEDULER_RECV_INTERVAL"
+    --stream-interval 50
+    "${TOKENIZER_ARGS[@]}"
+    --tokenizer-path "$MODEL"
+    --reasoning-parser qwen3
+    --tool-call-parser qwen3_coder
+    --speculative-algorithm EAGLE
+    --speculative-num-steps 3
+    --speculative-eagle-topk 1
+    --speculative-num-draft-tokens 4
+    --enable-metrics
+    --enable-cache-report
+)
+
+printf '%q ' "${SGLANG_CMD[@]}" | tee "$RESULT_DIR/sglang_command.txt"
+printf '\n' | tee -a "$RESULT_DIR/sglang_command.txt"
+"${SGLANG_CMD[@]}" > "$SERVER_LOG" 2>&1 &
+SERVER_PID=$!
+
+wait_for_server_ready --port "$PORT" --server-log "$SERVER_LOG" --server-pid "$SERVER_PID"
+
+if [ "${EVAL_ONLY:-false}" = "true" ]; then
+    run_eval --port "$PORT"
+else
+    build_replay_cmd "$RESULT_DIR"
+    REPLAY_CMD+=" --apply-chat-template"
+    run_agentic_replay_and_write_outputs "$RESULT_DIR"
+fi
@@ -374,6 +374,21 @@ qwen3.5-fp4-mi355x-sglang-mtp:
       - { tp: 2, conc-start: 4, conc-end: 128, spec-decoding: mtp }
       - { tp: 4, conc-start: 4, conc-end: 16, spec-decoding: mtp }
 
+qwen3.5-fp4-mi355x-sglang-agentic-mtp:
+  image: lmsysorg/sglang-rocm:v0.5.17-rocm720-mi35x-20260811
+  model: amd/Qwen3.5-397B-A17B-MXFP4
+  model-prefix: qwen3.5
+  runner: cluster:mi355x-amds
+  precision: fp4
+  framework: sglang
+  multinode: false
+  scenarios:
+    agentic-coding:
+    - dram-utilization: 0.80
+      search-space:
+      - { tp: 2, ep: 2, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20] }
+      - { tp: 4, ep: 1, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20, 24, 28, 32, 40] }
+
 qwen3.5-fp4-mi355x-sglang-disagg:
   image: lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260523
   model: amd/Qwen3.5-397B-A17B-MXFP4

diff --git a/perf-changelog.yaml b/perf-changelog.yaml
@@ -5874,3 +5874,11 @@
   description:
     - "Add MiniMax-M3 MXFP8 AgentX on H200 with vLLM v0.27.1, EAGLE3 golden AL 2.78, resident TP8 c1/c2/c4/c6/c8/c10, and Mooncake DRAM offload c12/c14."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2565
+
+- config-keys:
+    - qwen3.5-fp4-mi355x-sglang-agentic-mtp
+  description:
+    - "Add Qwen3.5-397B-A17B MXFP4 MI355X SGLang AgentX with native EAGLE MTP and the committed golden synthetic acceptance length."
+    - "Cover the measured resident TP2/EP2 and TP4 Pareto ranges through their HBM capacity knees, with required SGLang metrics exports."
+    - "Use SGLang v0.5.17 and disable unstable AITER all-reduce fusion for TP2/EP2 EAGLE rank consistency."
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2562