Performance Benchmarking of ExistentialSpecializer

[rajbarik]

A benchmark for showing performance improvements using ExistentialSpecializer. This shows more than 20x speedup when ExistentialSpecializer is used.

This benchmark has been significantly modified from the original implementation available at https://github.com/raywenderlich/swift-algorithm-club/tree/master/Bucket%20Sort
and
our issue created at:
kodecocodes/swift-algorithm-club#863

[rajbarik]

@atrick @slavapestov @eeckstein Please see the perf benchmark for ExistentialSpecializer.

[atrick]

Thanks Raj!

I think it would be helpful if you add comments at the top of the benchmark source explaining why it is useful and how it exercises existential specialization.

[rajbarik]

@atrick please see the description here.

[rajbarik]

I think it would be helpful if you add comments at the top of the benchmark source explaining why it is useful and how it exercises existential specialization.

I just uploaded a new version with comment. Please see it.

[atrick]

Thanks.

[atrick]

@swift-ci benchmark

[atrick]

@swift-ci smoke test

[swift-ci]

Performance: -O

TEST	OLD	NEW	DELTA	RATIO
Regression
ObjectiveCBridgeStubFromNSDateRef	4080	4390	+7.6%	0.93x (?)
Improvement
FlattenListLoop	4710	4337	-7.9%	1.09x
Array2D	8112	7520	-7.3%	1.08x
RemoveWhereSwapInts	70	65	-7.1%	1.08x
MapReduce	426	397	-6.8%	1.07x (?)
MapReduceAnyCollection	426	398	-6.6%	1.07x (?)

Performance: -Osize

TEST	OLD	NEW	DELTA	RATIO
Improvement
DataAppendDataLargeToLarge	50800	36800	-27.6%	1.38x (?)

How to read the data

The tables contain differences in performance which are larger than 8% and differences in code size which are larger than 1%.

If you see any unexpected regressions, you should consider fixing the
regressions before you merge the PR.

Noise: Sometimes the performance results (not code size!) contain false
alarms. Unexpected regressions which are marked with '(?)' are probably noise.
If you see regressions which you cannot explain you can try to run the
benchmarks again. If regressions still show up, please consult with the
performance team (@eeckstein).

Hardware Overview

  Model Name: Mac Pro
  Model Identifier: MacPro6,1
  Processor Name: 12-Core Intel Xeon E5
  Processor Speed: 2.7 GHz
  Number of Processors: 1
  Total Number of Cores: 12
  L2 Cache (per Core): 256 KB
  L3 Cache: 30 MB
  Memory: 64 GB

[slavapestov]

This breaks the build on Linux. Please use the Swift stdlib's random APIs instead of arc4random_uniform().

[rajbarik]

I see. Let me fix it @slavapestov