Document how to run and interpret XRay benchmarks - #3
Open
jimsweeney wants to merge 4 commits into
Open
jimsweeney wants to merge 4 commits into
jimsweeney wants to merge 4 commits into
Conversation
tpn
changed the title
tpn
force-pushed
the
xray-lp-benchmarks
branch
from
September 24, 2026 22:19
ad26ebf to
ab46916
Compare
|
Important Review skippedAuto reviews are limited based on label configuration. 🏷️ Required labels (at least one) (1)
Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository: NVIDIA/cuPhoton/.coderabbit.yaml Review profile: CHILL Plan: Advanced Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
Collaborator
|
@coderabbitai full review |
|
docs/xray/LINEAR-PREDICTION-BENCHMARKS.md records how lpb, lppb and lpsb measure (transfers excluded, warm-up call, null-stream synchronisation, best of --repeat), the reproduction commands, and the numbers on an RTX 4050 Laptop GPU under WSL2 with CUDA 13.2 and CuPy 14.1.1, together with a profile of where the P1 GPU time goes: cusolver xgeev is called once per 71 x 71 companion matrix even on the batched path, so batching buys 2 to 10 percent and the CPU is 15x faster at 96 samples. Linked from the xray README. No code changes. Refs NVIDIA#1 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014wyj5iPSHmtR3vSTvJWKfw Signed-off-by: James Sweeney <jmsweene@gmail.com>
Describe row variation, CPU batching and timed scalar transfers. State the missing component-order and profiling provenance without changing the contributed timing observations. Signed-off-by: Trent Nelson <trentn@nvidia.com>
Explain timing boundaries, CPU and GPU comparisons, and numerical checks for the existing benchmark commands. Keep machine measurements as review evidence and describe the versioned eigensolver limitation. Signed-off-by: Trent Nelson <trentn@nvidia.com>
tpn
force-pushed
the
xray-lp-benchmarks
branch
from
September 25, 2026 05:22
ab46916 to
59fcc31
Compare
Signed-off-by: Trent Nelson <trentn@nvidia.com>
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Documents how to run and interpret the existing XRay P1, P2, and Savitzky–Golay benchmarks. The guide explains the workloads, timing boundaries, JSON fields, CPU/GPU ratio denominators, numerical checks, and environment provenance needed to assess a run. It links the versioned CuPy eigensolver limitation and identifies matrix-pencil/ESPRIT as experimental estimator alternatives. Benchmark implementations and defaults are unchanged.
Validation on current public main: all three documented CPU variants and both subspace help commands passed; 12 documentation/schema tests and all applicable pre-commit hooks passed. The stage timing descriptions and JSON names match the implementation. No GPU execution was needed for this documentation refresh.
Historical contributor measurements
The original measurements and profiling observations remain in the original benchmark note. They describe revision
9e91835on one RTX 4050 Laptop GPU under WSL2. The exact profiling script/raw outputs were not retained, and the original P1 component count is uncertain. They are historical review evidence, not release performance baselines.Refs #1.