Merge origin/main into feat/measured-power-multinode

Resolve perf-changelog.yaml append conflict by keeping all three new
entries: main's #1579 (qwen3.5-fp4-mi355x-sglang-disagg) plus this
branch's #1574 re-trigger and the AMD multinode measured-power entry.
Append-only file (process_changelog rejects deletions); no lines removed.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

Author: arygupt

.github/configs/amd-master.yaml

@@ -494,6 +494,60 @@ qwen3.5-fp4-mi355x-sglang-mtp:
       - { tp: 2, conc-start: 4, conc-end: 256, spec-decoding: mtp }
       - { tp: 4, conc-start: 4, conc-end: 16, spec-decoding: mtp }
 
+qwen3.5-fp4-mi355x-sglang-disagg:
+  image: lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260523
+  model: amd/Qwen3.5-397B-A17B-MXFP4
+  model-prefix: qwen3.5
+  runner: mi355x-disagg
+  precision: fp4
+  framework: sglang-disagg
+  multinode: true
+  disagg: true
+  scenarios:
+    fixed-seq-len:
+    - isl: 1024
+      osl: 1024
+      search-space:
+      # 1P1D TP8/EP1, dp-attn false; MoRI conn.py overlay via job.slurm.
+      - spec-decoding: "none"
+        conc-list: [ 8, 16, 32, 64, 128, 256, 512 ]
+        prefill:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+          additional-settings:
+          - "PREFILL_NODES=1"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+          additional-settings:
+          - "DECODE_NODES=1"
+          - "DECODE_MTP_SIZE=0"
+
+    - isl: 8192
+      osl: 1024
+      search-space:
+      - spec-decoding: "none"
+        conc-list: [ 8, 16, 32, 64, 128, 256, 512 ]
+        prefill:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+          additional-settings:
+          - "PREFILL_NODES=1"
+        decode:
+          num-worker: 1
+          tp: 8
+          ep: 1
+          dp-attn: false
+          additional-settings:
+          - "DECODE_NODES=1"
+          - "DECODE_MTP_SIZE=0"
+
 qwen3.5-fp8-mi300x-sglang:
   image: lmsysorg/sglang:v0.5.12-rocm720-mi30x
   model: Qwen/Qwen3.5-397B-A17B-FP8

benchmarks/multi_node/amd_utils/models.yaml

@@ -161,6 +161,37 @@ DeepSeek-R1-0528:
       chunked_prefill_size: 262144
       cuda_graph_bs_range: "1-128"
 
+Qwen3.5-397B-A17B-MXFP4:
+  base_flags: "--decode-log-interval 1000 --log-level warning --watchdog-timeout 3600 --load-balance-method round_robin --kv-cache-dtype fp8_e4m3 --attention-backend aiter --disaggregation-transfer-backend mori --moe-dense-tp-size 1"
+  mtp_flags: ""
+  dp_flags: "--moe-a2a-backend mori --enable-dp-attention --enable-dp-lm-head"
+  prefill:
+    mem_fraction_static: 0.8
+    disable_radix_cache: true
+    dp:
+      max_running_requests: 24
+      chunked_prefill_size: "MORI_MAX_DISPATCH_TOKENS_PREFILL * PREFILL_TP_SIZE"
+      cuda_graph_bs: "1 2 3"
+    no_dp:
+      max_running_requests: 128
+      chunked_prefill_size: 262144
+      cuda_graph_bs_range: "1-128"
+  decode:
+    mem_fraction_static: 0.85
+    prefill_round_robin_balance: true
+    dp:
+      max_running_requests: 4096
+      chunked_prefill_size: "MORI_MAX_DISPATCH_TOKENS_DECODE * DECODE_TP_SIZE"
+      cuda_graph_bs_range: "1-160"
+    ep_only:
+      max_running_requests: 256
+      chunked_prefill_size: 262144
+      cuda_graph_bs_range: "1-256"
+    no_dp:
+      max_running_requests: 128
+      chunked_prefill_size: 262144
+      cuda_graph_bs_range: "1-128"
+
 Qwen3.5-397B-A17B-FP8:
   base_flags: "--decode-log-interval 1000 --log-level warning --watchdog-timeout 3600 --load-balance-method round_robin --kv-cache-dtype fp8_e4m3 --attention-backend aiter --disaggregation-transfer-backend mori --moe-dense-tp-size 1"
   mtp_flags: ""

benchmarks/multi_node/qwen3.5_fp4_mi355x_sglang-disagg.sh

@@ -0,0 +1,84 @@
+#!/usr/bin/env bash
+
+source "$(dirname "$0")/../benchmark_lib.sh"
+
+check_env_vars \
+    CONC_LIST \
+    ISL \
+    OSL \
+    IMAGE \
+    SPEC_DECODING \
+    MODEL_PATH \
+    PREFILL_NUM_WORKERS \
+    PREFILL_TP \
+    PREFILL_EP \
+    PREFILL_DP_ATTN \
+    DECODE_NUM_WORKERS \
+    DECODE_TP \
+    DECODE_EP \
+    DECODE_DP_ATTN \
+    PREFILL_NODES \
+    DECODE_NODES \
+    RANDOM_RANGE_RATIO \
+    FRAMEWORK
+
+if [[ -n "$SLURM_JOB_ID" ]]; then
+  echo "JOB $SLURM_JOB_ID running on $SLURMD_NODENAME"
+fi
+
+set -x
+
+# Use upstreamed multi_node scripts (no external clone needed)
+cd "$GITHUB_WORKSPACE/benchmarks/multi_node/amd_utils" || exit 1
+
+# Set up SGL launch script-specific environment variables
+export TIME_LIMIT="08:00:00"
+export MODEL_PATH=$MODEL_PATH
+export MODEL_NAME=$MODEL_NAME
+export CONTAINER_IMAGE=$IMAGE
+
+if [[ "${PREFILL_EP:-1}" -eq 1 ]]; then
+export PREFILL_ENABLE_EP=false
+else
+export PREFILL_ENABLE_EP=true
+fi
+
+if [[ "$PREFILL_DP_ATTN" == "true" ]]; then
+export PREFILL_ENABLE_DP=true
+else
+export PREFILL_ENABLE_DP=false
+fi
+
+if [[ "${DECODE_EP:-1}" -eq 1 ]]; then
+export DECODE_ENABLE_EP=false
+else
+export DECODE_ENABLE_EP=true
+fi
+
+if [[ "$DECODE_DP_ATTN" == "true" ]]; then
+export DECODE_ENABLE_DP=true
+else
+export DECODE_ENABLE_DP=false
+fi
+
+# Launch jobs based on ISL/OSL
+# Replace ' ' in CONC_LIST with 'x' such that the concurrency list is represented
+# by a list of numbers delimited by 'x'. This is because of how the underlying launch script
+# expects the concurrencies.
+JOB_ID=$(bash ./submit.sh $PREFILL_NODES \
+    $PREFILL_NUM_WORKERS \
+    $DECODE_NODES \
+    $DECODE_NUM_WORKERS \
+    $ISL $OSL "${CONC_LIST// /x}" inf \
+    ${PREFILL_ENABLE_EP} ${PREFILL_ENABLE_DP} \
+    ${DECODE_ENABLE_EP} ${DECODE_ENABLE_DP} \
+    ${PREFILL_TP} ${DECODE_TP} \
+    ${RANDOM_RANGE_RATIO} \
+    ${NODE_LIST:-})
+
+if [[ $? -ne 0 ]]; then
+    echo "Failed to submit job" >&2
+    exit 1
+fi
+
+echo "$JOB_ID"

perf-changelog.yaml

@@ -3213,3 +3213,11 @@
     - "The AMD amd_utils SLURM job has no orchestrator perfmon, so each SGLang/vLLM disagg node starts its own amd-smi monitor via start_perf_monitor (benchmarks/benchmark_lib.sh), writing perf_samples_<role>_w<idx>_<host>.csv into the NFS-shared /benchmark_logs/perfmon mount (wired in amd_utils/job.slurm). launch_mi355x-amds.sh collects the per-node CSVs into the GH workspace before the EXIT trap wipes the logs dir and sets GPU_METRICS_CSV_GLOB so the existing Process-result step runs the same vendor-agnostic utils/aggregate_power.py used for NVIDIA: per-source GPU-id namespacing (8 GPUs/node on MI355X, so a TP16 worker over 2 nodes counts 16 GPUs not 8), per-stage prefill/decode energy attribution, and per-worker temp/util/mem when amd-smi exposes those columns."
     - "Covers both engine paths: SGLang disagg (server_sglang.sh role = NODE_RANK bucketed by PREFILL_NODES_PER_WORKER / NODE_OFFSET) and vLLM disagg (server_vllm.sh one worker per node, ranks [0,xP) prefill / [xP,xP+yD) decode). Monitoring is best-effort end-to-end — a missing amd-smi or empty CSV skips power patching without failing the benchmark upload; DISAGG=true threads through to per-stage attribution while agg/non-disagg runs still get cluster-wide power."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1574
+
+- config-keys:
+    - qwen3.5-fp4-mi355x-sglang-disagg
+  description:
+    - "Add Qwen3.5-397B-A17B-MXFP4 MI355X SGLang PD-disaggregation"
+    - "Bump image to lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260523, 1P1D TP8/EP1, dp-attn false, conc [8..512]"
+    - "MoRI conn.py overlay (48e459bd) via job.slurm; launcher qwen3.5_fp4_mi355x_sglang-disagg.sh"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1579