Completely Unscientific Benchmarks

There are three kinds of lies: lies, damned lies, and statistics.

For this benchmark we implemented Treap in a few classic (C++, Java, Python) and hyped (JavaScript, Kotlin, Swift, Rust) programming languages and tested their performance on Linux, Mac OS, and Windows (all of them running on different hardware, so the results should not be compared between platforms).

This turned out to be a good benchmark of memory-intensive operations, which should have been pushed memory management implementations to their edge.

First, we tried to play by the rules of the garbage-collected languages, thus there are "ref-counted" versions of implementations for C++ and Rust, but then we still wanted to compare the results with idiomatic (a.k.a. common practices) implementations for C++ ("raw-pointers") and Rust ("idiomatic").

I must say that all the implementations except for C++ were implemented by mostly adapting the syntax from the very first implementation of the algorithm in Kotlin. Even Rust, which is considered to have the steepest learning curve among the tested languages, didn't require any "black magic" (the solution does not require either unsafe code or lifetime annotations). C++ was implemented separately, so it has a few shortcuts, and thus it might be not a completely fair comparison (I will try to implement "fair" C++ solution and also "C++"-like Rust solution to see if the performance can be on par).

Measurements

To measure time we used time util on Mac OS and Windows (msys2 environment), and cgmemtime on Linux.

Memory measurement was only available on Linux with cgmemtime util, which leverages CGroup capabilities to capture the high-water RSS+CACHE memory usage.

Results

Linux (Arch Linux, x64, Intel Core i7-4710HQ CPU)

Language	Real Time, seconds	Slowdown Time	Memory, MB	Normalized Memory	Binary Size, MB	Compiler Version
C++ "raw-pointers" (clang / gcc)	0.21	x1	0.38	x1	0.011 + libstdc++	Clang 6.0.0 / GCC 8.1.0
C++ "unique_ptr" (clang / gcc)	0.26	x1.24	0.38	x1	0.011 + libstdc++	Clang 6.0.0 / GCC 8.1.0
C++ "shared_ptr" ("ref-counted")	0.38	x1.8	0.5	x1.3	0.015 + libstdc++	Clang 6.0.0 / GCC 8.1.0
Rust "idiomatic"	0.37	x1.8	0.5	x1.3	0.427	Rustc 1.26
Rust "ref-counted"	0.37	x1.8	0.5	x1.3	0.431	Rustc 1.26
JavaScript	1.12	x5.3	52	x137	N/A	Node.js 10.1.0
Java (no-limit / -Xm*50M)	0.50 / 0.50	x2.4	142 / 29	x374 / x76	N/A	OpenJDK 1.8.0
Kotlin JVM (no-limit / -Xm*50M)	0.53 / 0.51	x2.5	144 / 30	x379 / x79	N/A	Kotlinc 1.2.40 + OpenJDK 1.8.0
Kotlin Native	5.88	x28	1.2	x3.2	0.239	Kotlinc-native 0.7
Swift	2.04	x9.7	2.5	x6.6	0.020 + Swift shared libraries	Swift 4.1
Nim	1.00	x4.8	0.5	x1.3	0.051	Nim 0.18 / GCC 8.1.0
Nim (gc:markAndSweep)	0.64	x3	5	x13	0.055	Nim 0.18 / GCC 8.1.0
Python (CPython)	12.25	x58.3	5	x13	N/A	CPython 3.6
Python (PyPy)	3.20	x15.2	48.5	x128	N/A	PyPy 6.0.0
C#	0.82*	x3.9	11	x29	N/A	.NET Core 2.0
Go	3.68	x17.5	8.6	x23	1.2	Go 1.10.2
D	0.24	x1.1	1.6	x4.2	0.019 + D runtime	LDC 1.9.0
Haskell	1.10	x5.2	3.4	x9	3.8	GHC 8.2.2

(*) C# has a noticable VM start time (~0.4 seconds), but we still measure real execution time of the whole program.

Mac OS (Mac OS 10.13, Intel Core i7-4770HQ CPU)

Language	Real Time, seconds	Slowdown Time	Binary Size, MB	Compiler version
C++ "raw-pointers" (clang)	0.25	x1	0.009 + libstdc++	Apple LLVM version 9.1.0 (clang-902.0.39.1)
C++ "shared_ptr" (clang)	1.35	x5.4	0.019 + libstdc++	Apple LLVM version 9.1.0 (clang-902.0.39.1)
Rust "ref-counted"	(needs update)	...	...	Rustc 1.26.0
Rust "idiomatic"	(needs update)	...	...	Rustc 1.26.0
JavaScript	1.47	x5.9	N/A	Node.js 6.11.1
Java (no-limit / -Xm*50M)	0.69 / 0.59	x2.8 / x2.4	N/A	Oracle JDK 1.8.0_131
Kotlin JVM (no-limit / -Xm*50M)	0.69 / 0.62	x2.8 / x2.5	N/A	Kotlinc 1.2.41 + Oracle JDK 1.8.0_131
Kotlin Native	8.2	x32.8	0.543	Kotlinc-native 0.6.2
Swift	2.2	x8.8	0.019 + Swift shared libraries	Apple Swift version 4.1
Nim	(needs update)	...	...	Nim 0.18
Python (CPython)	15.9	x63.6	N/A	CPython 2.7.10
Python (PyPy)	3.7	x14.8	N/A	PyPy 6.0.0

Windows (Windows 10, x64, Intel Core i7-6700HQ CPU)

Language	Real Time, seconds	Slowdown Time	Binary Size, MB	Compiler version
C++ "raw-pointers" (msvc 2017)	0.3	x1	0.015 + libstdc++	MSVC 2017 (19.13.26129)
C++ "shared_ptr" (msvc 2017)	1.7	x5.7	0.021 + libstdc++	MSVC 2017 (19.13.26129)
C++ "raw-pointers" (clang)	0.3	x1	0.254 + libstdc++	Clang 6.0.0
C++ "shared_ptr" (clang)	1.6	x5.3	0.258 + libstdc++	Clang 6.0.0
C++ "raw-pointers" (mingw)	1	x3.3	0.039 + libstdc++	GCC 6.3.0
C++ "shared_ptr" (mingw)	5	x16.7	0.031 + libstdc++	GCC 6.3.0
Rust "ref-counted"	(needs update)	...	...	Rustc 1.26.0
Rust "idiomatic"	(needs update)	...	...	Rustc 1.26.0
JavaScript	1.25	x4.2	N/A	Node.js 8.11.1
Java (no-limit / -Xm*50M)	0.8 / 0.75	x2.7 / x2.5	N/A	Oracle JDK 10.0.1
Kotlin JVM (no-limit / -Xm*50M)	0.8 / 0.8	x2.7 / x2.7	N/A	Kotlinc 1.2.41 + Oracle JDK 10.0.1
Kotlin Native	7.8	x26	0.46	Kotlinc-native 0.7
Swift (Swift for Windows)	2.5	x8.3	0.019 + Swift shared libraries	Swift 4.0.3 (Swift for Windows 1.9.1)
Nim	(needs update)	...	...	Nim 0.18
Python (CPython)	15.4	x51.3	N/A	CPython 2.7.13
Python (PyPy)	3.4	x11.3	N/A	PyPy 6.0.0

Observations

C++ "ref-counted" (shared ptr) has significant performance hit on non-Linux platforms.

JVM speeds up if you limit its memory.

JVM uses some tricks (JIT) which helps it to cut down some reference counting overheads and it manages to go faster than C++ and Rust "ref-counted" solutions.

Kotlin Native is still much slower than the Kotlin running in JVM.

Kotlin JS produces JS code which is ~25% slower than the manual Kotlin to JS translation.

With CPython vs PyPy you trade speed for memory.

License

Completely Unscientific Benchmarks project is licensed under either of

• Apache License, Version 2.0, (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
• MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)

at your option.