README.md

---
license: apache-2.0
tags:
  - profiling
  - ray
  - gpu
  - nsight-systems
  - ray-data
pretty_name: Ray Data GPU Idle Profiles (B200)
---

# Ray Data GPU Idle Profiles (B200)

Nsight Systems profiles (exported to SQLite, readable by
[`nsys-ai`](https://github.com/GindaChen/nsys-ai)) from an experiment on how a
Ray Data pipeline keeps a GPU idle, and how the loss splits between *moving
data* and *waiting for data*. Captured on a single NVIDIA B200 with Ray 2.58.0
/ master, PyTorch 2.14.0+cu130, Nsight Systems 2026.1.3.

These profiles back the write-up in the iThome Ironman series
「GPU 很忙？他真的有在做事嗎？」 (Days 27–29), and are shared so the numbers and
the before/after comparison are independently reproducible.

## The workload

A deliberately small pipeline so scheduling shows through instead of the model:

```
ray.data.range → 32 blocks → 2048×8192 FP32 (64 MiB) per batch
  → H2D → 9 matmuls (~39 ms) → D2H;  CPU prep ~40 ms per batch
```

Two knobs move across the files: where data prep runs (inside the GPU UDF vs a
separate CPU operator), and whether pinned staging overlaps with compute.

## Files

Analysis window is each profile's kernel span on `device 0`
(`CUDA_VISIBLE_DEVICES` exposes one physical card, renumbered to 0).

| File | Stage | device idle | copy_ms | note |
| --- | --- | --- | --- | --- |
| `inline.sqlite` | prep inside the GPU UDF | 59.0% | — | serial load→compute |
| `split.sqlite` | prep as a CPU operator | 29.1% | — | H2D 8.5 GB/s pageable |
| `pinned.sqlite` | naive `pin_memory` in UDF | 31.4% | — | worse: staging still serial |
| `overlap.sqlite` | `enable_true_multi_threading` + per-thread stream | 24.7% | — | 31.3% of H2D overlapped |
| `patched_split.sqlite` | clean profile after fixing Ray nsight bug | 31.3% | 339.6 | idle split by copy vs wait |
| `final_split.sqlite` | **baseline** for the prototype | 62.4% | 830.3 | before pinned-staging |
| `final_staged.sqlite` | **after** map_batches pinned-staging prototype | 54.8% | 155.1 | copy_ms −82% |

`patched_split.nsys-rep` is the raw report for opening in the Nsight Systems GUI.

Code:

- `workload.py` — the four modes (inline / split / pinned / overlap).
- `workload_staged.py` — the `ActorPoolStrategy(pinned_staging=True)` prototype run.
- `repro_flush.py` — minimal A/B for the nsys teardown-flush race
  ([ray#60904](https://github.com/ray-project/ray/issues/60904)).
- `our_changes.patch` — the map_batches pinned-staging prototype (vs Ray master `c8466ab8`).
- `cpp_fix_v3.diff` — the C++ fix for the teardown-flush race
  ([ray#66129](https://github.com/ray-project/ray/pull/66129)).

## Reproduce the analysis

```bash
pip install nsys-ai
# the copy_ms field (newer nsys-ai) splits "moving data" out of the idle
nsys-ai skill run gpu_idle_gaps final_split.sqlite  -p device=0 --format json
nsys-ai skill run gpu_idle_gaps final_staged.sqlite -p device=0 --format json
```

## Two Ray bugs found while capturing these

Getting a trustworthy profile out of Ray Data first meant fixing two Ray Core
bugs, both reduced to minimal repros and reported upstream:

- **nsight runtime_env drops the selected Python command** (worker never starts):
  [ray#66093](https://github.com/ray-project/ray/issues/66093) /
  [PR #66094](https://github.com/ray-project/ray/pull/66094).
- **nsys report killed during teardown** (empty/missing `.nsys-rep`):
  [ray#60904](https://github.com/ray-project/ray/issues/60904) /
  [PR #66129](https://github.com/ray-project/ray/pull/66129).

## Caveats

Numbers with the profiler on run larger than without it; the paired *difference*
is what matters, not the absolute idle. The box was shared, so wall-time
comparisons use the minimum across runs. See the series notes for the full
methodology and the "local metric vs mechanism vs end-to-end" distinction.