Add PrecompileTools workload to OptimizationBase for improved TTFX #1119
+26
−0
Conversation
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
This adds a precompilation workload to OptimizationBase that precompiles the most common operations: OptimizationFunction and OptimizationProblem creation with and without bounds. ## Performance improvements - TTFX for OptimizationProblem creation: 0.081s -> 0.024s (~70% improvement) - Loading time remains essentially unchanged (~1.33s) ## Analysis performed - Measured baseline startup and TTFX times - Checked for invalidations using SnoopCompile (32 minor trees, all from dependencies, none actionable) - Profiled inference during TTFX to identify precompilation targets - Added minimal precompile workload for common use cases 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <[email protected]>
Member
Those types are in SciMLBase, so the precompilation here is in the wrong repo. |
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Outdated suggestions cannot be applied.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.
Summary
Performance Improvements
Analysis Performed
Baseline measurements: Measured startup time (~1.33s) and TTFX for OptimizationProblem creation (~0.081s)
Invalidation analysis: Used SnoopCompile to check for invalidations. Found 32 minor invalidation trees (1 child each), all from external dependencies (FillArrays, SparseConnectivityTracer, RecursiveArrayTools, SciMLBase) - none actionable for this package.
TTFX profiling: Used @snoop_inference to profile inference during first use. Most time spent in Base, Core, and SparseMatrixColorings rather than OptimizationBase/SciMLBase code.
Precompilation workload: Added minimal workload that covers the most common use case - creating OptimizationFunction and OptimizationProblem with and without bounds.
Test Results
All OptimizationBase tests pass (703 tests in ~8.5 minutes).
Changes
lib/OptimizationBase/Project.toml: Added PrecompileTools dependencylib/OptimizationBase/src/OptimizationBase.jl: Include precompilation filelib/OptimizationBase/src/precompilation.jl: New precompilation workloadcc @ChrisRackauckas
🤖 Generated with Claude Code