swift-vs-python

Non-scientific, somewhat real-world benchmark to evaluate Swift vs Python performance. Now expanding to include more common programming languages

Standard test configuration: Using the same language-specific syntax/flow control...Increment an array of 10 integers 10 million times each.

How to use

Run benchmark.sh to run the benchmarks.

Preliminary Results

Swift's compiler seems to optimize this iterative increment operation somehow, resulting in a near constant time to run this benchmark.

Meanwhile, Python is unable to finish in a reasonable amount of time...

Example Results

10 Million increments each array item, Python takes too long for more, and Java doesn't support integers larger than that.

sam$ ./benchmark.sh 

Running Swift Benchmark...
real	0m0.485s
user	0m0.429s
sys	0m0.046s
Swift Output:
[10000001, 10000002, 10000003, 10000004, 10000005, 10000006, 10000007, 10000008, 10000009, 10000000]

================================
Running Swift Multithreaded Benchmark...
real	0m0.897s
user	0m0.695s
sys	0m0.189s
Swift Output:
[10000001, 10000002, 10000003, 10000004, 10000005, 10000006, 10000007, 10000008, 10000009, 10000000]

================================
Running Java Benchmark...
real	0m0.725s
user	0m1.351s
sys	0m0.106s
Java Output:
[10000001, 10000002, 10000003, 10000004, 10000005, 10000006, 10000007, 10000008, 10000009, 10000000]

================================
Running Python Benchmark...
real	0m14.475s
user	0m14.427s
sys	0m0.034s
Python Output:
[10000001, 10000002, 10000003, 10000004, 10000005, 10000006, 10000007, 10000008, 10000009, 10000000]

================================
Running C++ Benchmark...
real	0m0.930s
user	0m0.743s
sys	0m0.066s
C++ Output:
[10000001, 10000002, 10000003, 10000004, 10000005, 10000006, 10000007, 10000008, 10000009, 10000000]