Issues

• [PowerX] centralize DCGM profiles and backfill power coverage / 集中配置 DCGM profiles 并补齐功耗覆盖

  SemiAnalysisAI/InferenceX#3592 · Filter by author edwingao28 opened on Sep 29, 2026

• B300 multi-node Dynamo TRT-LLM is broken on DSXE

  SemiAnalysisAI/InferenceX#3425 · Filter by author cquil11 opened on Sep 25, 2026

• [AgentX GB200] restore the Kimi-K3 18-point arm: telemetry gate on 8 completed points, engine failures on 8 / 恢复 GB200 Kimi-K3 的 18 点 arm：8 个跑完的点被功耗闸门拦下，8 个在引擎内失败

  SemiAnalysisAI/InferenceX#3322 · Filter by author edwingao28 opened on Sep 20, 2026

• [Question] Is hardware contribution to the official CI cluster supported?

  SemiAnalysisAI/InferenceX#3122 · Filter by author liuming10 opened on Sep 15, 2026
  3

• Runtime JIT compilation contaminates published TTFT and TPOT measurements

  SemiAnalysisAI/InferenceX#2819 · Filter by author simone-chen opened on Sep 3, 2026

• Single-node num_gpus ignores --data-parallel-size, inflating tput_per_gpu

  SemiAnalysisAI/InferenceX#2714 · Filter by author aistackdev opened on Aug 24, 2026

• [Power] Power measurement roadmap

  SemiAnalysisAI/InferenceX#2681 · Filter by author edwingao28 opened on Aug 19, 2026
  

• [CI] Recipe bring-up for /overview combos with no data / [CI] 缺数据 combo 的 recipe bring-up 跟踪

  SemiAnalysisAI/InferenceX#2587 · Filter by author edwingao28 opened on Aug 12, 2026

• [WIP] refactor/cleanup repo

  SemiAnalysisAI/InferenceX#2382 · Filter by author cquil11 opened on Jul 28, 2026
  

• Proposal: add Unofficial DeepSeek-V4-Flash 4xH200 SGLang recipe / 提案：新增非官方 DeepSeek-V4-Flash 4×H200 SGLang 配方

  SemiAnalysisAI/InferenceX#2315 · Filter by author kfastino opened on Jul 23, 2026
  2

• [CI] Same-batch comparability for /overview data / /overview 数据的同批次可比性

  SemiAnalysisAI/InferenceX#2304 · Filter by author edwingao28 opened on Jul 22, 2026
  

• 请问是否有长序列的比较维度？

  SemiAnalysisAI/InferenceX#2287 · Filter by author ghostplant opened on Jul 20, 2026
  12

• measuring power

  SemiAnalysisAI/InferenceX#2250 · Filter by author functionstackx opened on Jul 16, 2026

• Upstream SGLang fix for disaggregated prefill TP-rank desync / 向 SGLang 上游提交分离式预填充 TP rank 不同步修复

  SemiAnalysisAI/InferenceX#2151 · Filter by author cquil11 opened on Jul 10, 2026

• Clean up configuration ownership and redundant defaults after PR #2127 / 清理 PR #2127 后的配置归属与冗余默认值

  SemiAnalysisAI/InferenceX#2150 · Filter by author cquil11 opened on Jul 10, 2026

• Removal of "1 The ATOM engine is promising, however it has yet to serve production tokens. It is still in its infant stage."

  SemiAnalysisAI/InferenceX#2125 · Filter by author seungrokj opened on Jul 9, 2026
  1

• Refactor common result metadata processing into a shared helper

  SemiAnalysisAI/InferenceX#2105 · Filter by author cquil11 opened on Jul 6, 2026
  1

  

• Add gate on full-sweep / fail-fast runs to prevent wasted compute

  dev velocityenhancementp1
  SemiAnalysisAI/InferenceX#2103 · Filter by author cquil11 opened on Jul 6, 2026
  

• runners.yaml: add per-cluster directory mappings (centralize paths hardcoded in launch_*.sh) / runners.yaml：添加集群目录映射（集中管理 launch_*.sh 中硬编码的路径）

  SemiAnalysisAI/InferenceX#1973 · Filter by author cquil11 opened on Jul 1, 2026
  

• Add local "dry run" mode: run all GHA input/script-generation logic without triggering a run / 添加本地「试运行」模式：执行所有 GHA 输入/脚本生成逻辑而不触发实际运行

  SemiAnalysisAI/InferenceX#1972 · Filter by author cquil11 opened on Jul 1, 2026
  

• Generalize per-subpoint benchmark checkpointing for reruns / 泛化基准测试子点检查点机制以支持重跑

  SemiAnalysisAI/InferenceX#1971 · Filter by author cquil11 opened on Jul 1, 2026
  

• Template the benchmark scripts / 模板化基准测试脚本

  SemiAnalysisAI/InferenceX#1945 · Filter by author cquil11 opened on Jun 26, 2026
  

• Quality Evals that stress longer decode / 压力测试长解码的质量评估

  SemiAnalysisAI/InferenceX#1920 · Filter by author sshleifer opened on Jun 24, 2026
  2

• Front End: Select Date Range too often / 前端：日期范围选择过于频繁

  SemiAnalysisAI/InferenceX#1919 · Filter by author sshleifer opened on Jun 24, 2026
  1

• [enhancement] add more info to runners yaml / [enhancement] 在 runners yaml 中添加更多信息

  SemiAnalysisAI/InferenceX#1910 · Filter by author cquil11 opened on Jun 23, 2026