Skip to content

Ascend/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.pyscript9.0 KB
scripts/delta_report.pyscript4.8 KB
scripts/env_check.shscript5.0 KB
scripts/result_saver.pyscript5.0 KB

references/cases/tilelang/pr_tilelang_persistent_mqa_decode.md

View on GitHub
# PR #1390 "perf(HISA): 使用 persistent 调度优化 A3 paged block-sparse MQA decode"【PR】 数据不严谨声明

## 基本信息
- 算子类别:attention
- DSL/框架:tilelang
- 类型:PR
- 来源可信度:一手 PR 原文( 但 PR 作者注明数据来自非标准不严谨测试,见"适用范围与警示")

## 来源链接
- PR/出处链接:<https://github.com/tile-ai/tilelang-ascend/pull/1390>( 已验证可达,验证日期 2026-08;2026-07-16 提交,抓取时为 open 状态)
- 优化代码查看:<https://github.com/tile-ai/tilelang-ascend/pull/1390/files>

## 问题与瓶颈
task-grid 调度下每个短 task 都有独立 mixed-block 初始化/流水排空/AIC-AIV 完成确认;当 AIC 已完成 MMA 而 AIV 仍在 reduce/mask/MTE3 输出时产生集群内 C/V 调度空洞。

## 优化方法(理论手段)
1. task-grid → persistent resident block(`T.Persistent`,physical_grid = min(logical_grid, aicore_num))——常驻 block 复用,摊销每 task 的初始化/排空开销。
2. 新增 A3 单发布窗口 V→C release 协议(`PERSISTENT_QUEUE_DEPTH=1`,release 放在 late reduction 后、tail mask 前)——使下一 task 的 AIC 预计算覆盖上一 task 的 AIV 尾部,消除 C/V 调度空洞。
3. 闭合 mode2 广播 ready event 生命周期——避免 event 泄漏/悬挂。
4. 显式 `T.wait_flag("M","MTE1",...)` 保护跨迭代 L1/L0 buffer——防止跨迭代数据竞争。
5. mask 改 full/invalid/partial 三路——热路径无 mask 指令。

## 性能对比
PR 正文数据(Ascend910_9362,20 AIC/40 AIV,1650MHz;seq_len=1, topk=64, heads=32, index_dim=128, kv_block=128):

| Batch | 原 task-grid | Persistent | 耗时降低 | 加速比 |
|---|---|---|---|---|
| 1 | 12.040 us | 10.940 us | 9.1% | 1.10x |
| 2 | 18.000 us | 13.620 us | 24.3% | 1.32x |
| 4 | 28.319 us | 18.700 us | 34.0% | 1.51x |
| 8 | 42.979 us | 27.839 us | 35.2% | 1.54x |

## 适用范围与警示
- 适用场景:A3 paged block-sparse MQA decode;测试 shape 为 seq_len=1, topk=64, heads=32, index_dim=128, kv_block=128。
- ** 风险说明**:PR 作者注明"该数据来自开发过程中的非标准不严谨测试",引用时需保留此说明。
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