Skip to content

kali20gakki/msAgent / skills/operator/ops-performance-tuning

ops-performance-tuning

Verified

Compile, profile, diagnose, optimize, and compare Ascend NPU operators across CANN versions and A2/A3/A5 for Ascend C, CATLASS, Triton-Ascend, TileLang-Ascend, PyPTO, and SHMEM/MC2. Use when an operator already has a runnable implementation and the user asks for msOpProf collection, bottleneck analysis, source-level tuning, or a reproducible before/after performance report. Do not use as the primary workflow for operator creation, migration, or unresolved correctness failures.

FileKindSize
SKILL.mdskill md15.4 KB
evals/evals.jsonconfig3.6 KB
references/case-routing.mdmarkdown30.0 KB
references/cases/ascendc/pr_ascendc_allgathermm_scalar_align.mdmarkdown1.7 KB
references/cases/ascendc/pr_ascendc_cann_samples_perf_story.mdmarkdown3.9 KB
references/cases/ascendc/pr_ascendc_cgdr_vf_inverse.mdmarkdown1.6 KB
references/cases/ascendc/pr_ascendc_conv3ddx_tque_tbuf.mdmarkdown1.6 KB
references/cases/ascendc/pr_ascendc_layernorm_areg.mdmarkdown1.6 KB
references/cases/ascendc/pr_ascendc_logit_a5_microapi.mdmarkdown1.9 KB
references/cases/ascendc/pr_ascendc_matmul_case_18x.mdmarkdown1.5 KB
references/cases/ascendc/pr_ascendc_matmul_full_load_wiki.mdmarkdown1.6 KB
references/cases/ascendc/pr_ascendc_matmul_vcv_wiki.mdmarkdown1.8 KB
references/cases/ascendc/pr_ascendc_mc2_fusion.mdmarkdown1.4 KB
references/cases/ascendc/pr_ascendc_mhc_24x.mdmarkdown1.2 KB
references/cases/ascendc/pr_ascendc_msopprof_add_block_imbalance.mdmarkdown1.7 KB
references/cases/ascendc/pr_ascendc_quantize_a2_micro_opt.mdmarkdown2.0 KB
references/cases/ascendc/pr_ascendc_rotate_quant_blaze.mdmarkdown2.8 KB
references/cases/ascendc/pr_ascendc_tiling_l2_official.mdmarkdown1.6 KB
references/cases/catlass/pr_catlass_baseline.mdmarkdown1.3 KB
references/cases/catlass/pr_catlass_fixpipe_dualdst.mdmarkdown1.6 KB
references/cases/catlass/pr_catlass_grouped_matmul_tiling.mdmarkdown1.5 KB
references/cases/catlass/pr_catlass_mstuner_3p8x.mdmarkdown1.9 KB
references/cases/catlass/pr_catlass_swizzle_balance.mdmarkdown1.6 KB
references/cases/pypto/pr_pypto_gqa_decode_vector_merge.mdmarkdown2.4 KB
references/cases/pypto/pr_pypto_l2_hit_layout_s13.mdmarkdown1.9 KB
references/cases/pypto/pr_pypto_per_matmul_tile_shapes.mdmarkdown2.4 KB
references/cases/pypto/pr_pypto_small_shape_matmul.mdmarkdown2.3 KB
references/cases/pypto/pr_pypto_weight_none_cacheable.mdmarkdown2.6 KB
references/cases/shmem/pr_shmem_alltoallv_mr152.mdmarkdown2.4 KB
references/cases/shmem/pr_shmem_reduce_scatter_serial.mdmarkdown2.6 KB
references/cases/tilelang/pr_tilelang_benchmark_vs_ascendc.mdmarkdown1.7 KB
references/cases/tilelang/pr_tilelang_compiler_prs.mdmarkdown2.3 KB
references/cases/tilelang/pr_tilelang_fa_pr698.mdmarkdown1.5 KB
references/cases/tilelang/pr_tilelang_gqa_bwd_pr1494.mdmarkdown1.3 KB
references/cases/tilelang/pr_tilelang_persistent_mqa_decode.mdmarkdown2.2 KB
references/cases/tilelang/pr_tilelang_sfa_pr665.mdmarkdown914 B
references/cases/triton/pr_triton_atomic_simd_24x.mdmarkdown1.8 KB
references/cases/triton/pr_triton_autotune_background.mdmarkdown1.0 KB
references/cases/triton/pr_triton_constexpr_recompile.mdmarkdown1.1 KB
references/cases/triton/pr_triton_lgamma_simt.mdmarkdown1.6 KB
references/compile/compile-ascendc.mdmarkdown9.0 KB
references/compile/compile-catlass.mdmarkdown2.4 KB
references/compile/compile-pypto.mdmarkdown2.0 KB
references/compile/compile-shmem.mdmarkdown2.6 KB
references/compile/compile-tilelang.mdmarkdown2.7 KB
references/compile/compile-triton.mdmarkdown4.4 KB
references/diagnose/diagnose-l2-cache.mdmarkdown2.2 KB
references/diagnose/diagnose-occupancy.mdmarkdown4.0 KB
references/diagnose/diagnose-pipeline.mdmarkdown4.1 KB
references/diagnose/diagnose-roofline.mdmarkdown3.1 KB
references/optimize/optimize-api-usage.mdmarkdown1.7 KB
references/optimize/optimize-ascendc-regbase-vf.mdmarkdown5.9 KB
references/optimize/optimize-ascendc-tiling.mdmarkdown935 B
references/optimize/optimize-catlass.mdmarkdown3.0 KB
references/optimize/optimize-data-copy.mdmarkdown870 B
references/optimize/optimize-index.mdmarkdown3.6 KB
references/optimize/optimize-memory-hierarchy.mdmarkdown599 B
references/optimize/optimize-pipeline.mdmarkdown1.0 KB
references/optimize/optimize-pypto.mdmarkdown1.5 KB
references/optimize/optimize-shmem.mdmarkdown5.1 KB
references/optimize/optimize-tilelang.mdmarkdown1.7 KB
references/optimize/optimize-triton.mdmarkdown3.8 KB
references/profile/benchmark-harness.mdmarkdown2.2 KB
references/profile/profile-msopprof.mdmarkdown7.4 KB
references/troubleshooting.mdmarkdown5.8 KB
scripts/bound_analyzer.pyscript9.8 KB
scripts/delta_report.pyscript5.5 KB
scripts/env_check.shscript5.7 KB
scripts/result_saver.pyscript5.8 KB

references/diagnose/diagnose-roofline.md

View on GitHub
# roofline

> roofline(§3):msOpProf `--aic-metrics=Roofline` 瓶颈分析与 bound 类型判读;工具总览与五类对照表见 [msOpProf 采集指南](../profile/profile-msopprof.md)。

## 1. 特性介绍

Roofline 瓶颈分析:把算子的实际算力/带宽投射到硬件理论上限的 roofline 图上,判定算子是 **Compute Bound / Memory Bound / Latency Bound**,从而决定优化方向(提算力利用率、减访存流量、还是消延迟/断档)。

## 2. 特性功能

- **输出文件**:`visualize_data.bin` 用 MindStudio Insight 看 Roofline 图;瓶颈数据源 CSV 含 `ArithmeticUtilization.csv`、`Memory.csv` 等。
- **判读结论**:bound 类型三分类(Compute / Memory / Latency Bound),与 Default 基础指标绑定使用。

## 3. 如何使用

- 命令行:若本机帮助支持,使用 msOpProf `Roofline` 并按该版本要求与 `Default` 组合。以位置参数形式为例(输出目录需先 `chmod 755`):

  ```bash
  mkdir -p msprof_roofline && chmod 755 msprof_roofline
  msprof op --output=./msprof_roofline --aic-metrics=Roofline \
    --kernel-name=add_custom ./execute_add_op
  ```

  参数形式和芯片支持边界见 [跨版本探测模板](case-routing.md#3-msopprof-跨版本采集先探测再选命令)。判读顺序:先看 Default 基础指标 → 开 Roofline 判 bound 类型。
- 使用场景:拿到基线后、动手优化前判定 bound 类型;也用于优化后确认是否已贴近硬件上限(继续调是否还有空间)。

无 Roofline 图时的**利用率式自建判据**:

- 通信算子 `utilization = bus_bw / peak_bw`;
- CATLASS "Cube 利用率 ~90%+ 即接近 roofline,再提速受硬件结构约束";
- MTE2 Bound 先判断是否已达理论带宽;
- Triton simulator 侧真·硬件极限判据:`MMAD`>50%。

## 4. 分析案例

若当前归档中没有 Roofline 专属输出,不得推断 Roofline Bound。先按上一节检查本机帮助并补采;版本不支持时改用 OpBasicInfo、PipeUtilization、带宽与算量共同给出 `UNRESOLVED` 或低置信度结论。

可作为替代证据的是 Roofline 瓶颈数据源 CSV(`ArithmeticUtilization.csv`、`Memory.csv`)在该用例上的真实 before/after 对照——它支撑了"Memory Bound(搬运方式)"的判定:

| 指标(字段) | before | after |
|---|---|---|
| `ArithmeticUtilization.csv` → `aiv_vec_ratio` | 0.0594~0.0660(Vec 仅约 6%) | 0.2041~0.2074(约 20.5%) |
| `Memory.csv` → `aiv_gm_to_ub_bw(GB/s)` | 2.031~2.258 | 70.16~71.29 |
| `Memory.csv` → `GM_to_UB_bw_usage_rate(%)` | 1.245~1.384 | 43.01~43.70 |

(均为 block_id 0–7 各行范围)

判读:before 时计算利用率约 6%、GM→UB 带宽使用率仅约 1.2~1.4%——算力与带宽都离 roofline 上限很远,属"搬运方式瓶颈"(400B 非对齐小包)而非真带宽上限;after 改 16KB 对齐大包 + Double Buffer 后带宽使用率升至约 43%、带宽 2.0→70.6 GB/s(约 33 倍),Vec 利用率同步升至约 20.5%,仍非 Compute Bound——即"瓶颈=搬运方式,不是带宽上限也不是计算"。
planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies.Stars: 108kUpdated Claude CodeCodexVerifiedInstall
meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要".Stars: 108kUpdated Claude CodeCodexVerifiedInstall