Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

2.x: benchmark the new strict/interop mode #5115

Merged
merged 2 commits into from
Feb 18, 2017
Merged

Conversation

akarnokd
Copy link
Member

Benchmark the overhead of the strict/interop mode.

i5 6440HQ, Windows 10 x64, Java 8u121

image

The numbers are consistent with my expectations; this mobile processor is roughly equivalent to i7 4770 desktop where the cost model is: 1 atomic increment per item equals to roughly 130 Mops/s upper limit, 2 atomic increment per item is roughly 60 Mops/s upper limit. Since the interop mode requires at minimum two atomic increments, 54 Mops/s is a reasonable value to get.

@akarnokd akarnokd added this to the 2.1 milestone Feb 18, 2017
@codecov
Copy link

codecov bot commented Feb 18, 2017

Codecov Report

Merging #5115 into 2.x will decrease coverage by -0.03%.
The diff coverage is n/a.

@@             Coverage Diff              @@
##                2.x    #5115      +/-   ##
============================================
- Coverage     95.15%   95.12%   -0.03%     
- Complexity     5561     5567       +6     
============================================
  Files           620      620              
  Lines         40011    40011              
  Branches       5632     5632              
============================================
- Hits          38072    38061      -11     
- Misses          920      939      +19     
+ Partials       1019     1011       -8
Impacted Files Coverage Δ Complexity Δ
...internal/disposables/ArrayCompositeDisposable.java 92.85% <ø> (-7.15%) 15% <ø> (-1%)
...ors/observable/ObservableSampleWithObservable.java 91.56% <ø> (-6.03%) 3% <ø> (ø)
...l/operators/observable/ObservableFlatMapMaybe.java 89.54% <ø> (-5.23%) 2% <ø> (ø)
...in/java/io/reactivex/subjects/BehaviorSubject.java 84.97% <ø> (-4.67%) 56% <ø> (+1%)
...ava/io/reactivex/processors/BehaviorProcessor.java 87.38% <ø> (-4.21%) 56% <ø> (+1%)
...io/reactivex/internal/util/BackpressureHelper.java 95.91% <ø> (-4.09%) 21% <ø> (-1%)
...ternal/operators/observable/ObservablePublish.java 91.89% <ø> (-3.61%) 10% <ø> (ø)
...al/operators/observable/ObservableSampleTimed.java 93.33% <ø> (-3.34%) 3% <ø> (ø)
.../operators/observable/ObservableFlatMapSingle.java 94.02% <ø> (-2.99%) 2% <ø> (ø)
...operators/observable/ObservableConcatMapEager.java 97.29% <ø> (-2.17%) 2% <ø> (ø)
... and 34 more

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a00ea07...95ed71a. Read the comment docs.

@akarnokd
Copy link
Member Author

Results with the updated benchmark, adding extra overhead by consuming CPU to see where the raw overhead decreases to a reasonably low value:

image

@akarnokd akarnokd merged commit 0518d58 into ReactiveX:2.x Feb 18, 2017
@akarnokd akarnokd deleted the StrictPerf branch March 24, 2017 13:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants