git:20260727.92007b7 to git:20260804.b3bf3e7

---

name: dfx-analyze

description: Analyze an onboard run's performance/scheduling/dependency/dump data using simpler's BUILT-IN DFX tools (simpler_setup.tools.*) instead of hand-rolling instrumentation. Use AFTER an onboard run when you need per-run device timing (Total/Orch/Sched), AICPU scheduler-overhead / Tail-OH breakdown, the task dependency graph, scope ring-fill peaks, or to inspect args dumps. These are simpler's own tools (shipped in the wheel), distinct from any cross-repo workload. Reach for this before writing custom timing/logging into the runtime.

---

# Analyze DFX data (simpler's own tools)

simpler already ships end-user analysis CLIs under `simpler_setup.tools` —

**use them; do not re-invent timing/instrumentation in the runtime.** Canonical

reference (tool flags, examples, output paths): `simpler_setup/tools/README.md`.

Per-DFX docs: `docs/dfx/` (`l2-timing.md`, `sched-overhead-model.md`,

## Pick the tool by question

| You want… | Tool | Needs |

| --------- | ---- | ----- |

| Per-run **Host / Device / Effective / Orch / Sched** timing | `strace_timing --rounds-table` | nothing extra — `[STRACE]` markers are on stderr (`SIMPLER_HOST_STRACE`, compile-time default on, **NOT** gated by swimlane) |

| Task **dependency graph** (text / HTML) | `deps_viewer` | `--enable-dep-gen` run → `deps.json` |

| **Per-scope ring-fill peaks** (task_window / heap / tensormap) | `scope_stats_plot` | `--enable-scope-stats` run → `scope_stats.jsonl` |

| Inspect / export **args dumps** | `dump_viewer` | `--dump-args` run → `args_dump/` |

## First reflex: Host/Device/Orch/Sched needs nothing extra

To answer "where did the time go / is this AICPU-orchestration bound", you do

**not** need swimlane or custom logging — just tee the run's stderr, then:

```bash

python test_*.py -p <platform> -d <device> --rounds N --skip-golden > run.log 2>&1

python -m simpler_setup.tools.strace_timing run.log --rounds-table

# prints per-round Host / Device / Effective / Orch / Sched (us);

# Orch≈Sched≈Effective ⇒ AICPU-bound. (Effective = orch∪sched window.)

# --tree instead shows the nested span tree (device_wall → preamble/so_load/

# graph_build → config_validate/arena_wire/sm_reset prep + orch/sched → post_orch).

# SIMPLER_DEVICE_STRACE_ENABLE=0 drops the device clk=dev markers (host spans stay).

```

For the per-thread `loops`/`tasks_scheduled` deep-dive (not in the markers),

rebuild with `SIMPLER_SCHED_PROFILING=1` and read the device log directly.

## Where the inputs are written

DFX artifacts land in the run's output dir with fixed filenames:

- simpler scene tests (`tests/st`): `outputs/<case>_<ts>/` (the tools auto-pick

the latest by mtime when run from the dir holding `outputs/`).

- JIT examples / pypto-lib: `build_output/_jit_*/dfx_outputs/`.

## Don't

- ❌ Hand-roll per-stage / submit-drain / per-scope timing in the runtime to get

numbers these tools already produce. If a tool is missing a metric, extend the

tool, not the hot path (and never log on AICPU hot paths — see

`codestyle.md` rule 7).