fastfuck

Benchmarking languages with Brainfuck

The what and the why

Brainfuck as a language is very similar to P``, one of the simplest turing complete languages. It also happens to be remarkably simple to write an interpreter for, and there are a vast library of preexisting brainfuck programs, so I thought it would be an interesting project to use as a language benchmark.

For each language benchmarked, there are 2 versions:

• idiomatic - this contains idiomatic, normal, clean code you'd expect in a professional environment, understandable by anyone familiar with the language
• optimised - this contains hyper-optimised code which only has the aim of being as fast as possible - clarity/correctness be damned

What this benchmark does do

• Shows the general order-of-magnitude performance of a languae
  • You can safely use this benchmark to determine that Ruby, is, in fact, slower than C
• Demonstrates the capabilities of a language in extremely monotonic data processing, such as text-parsing

What this benchmark doesn't do

• Determine the performance of a language in all real-world contexts
  • Notably, this benchmark will give a big advantage to systems which can JIT or otherwise optimise very hot methods, as the main loop of this program will be executed millions of times
• Determine the overally quality of a language
  • I hope this one needs no explanation

Can I add language?

Absolutely - contributions are welcome. The structure for adding a new language is extremely simple:

1. Add a new directory under languages, named appropriately
2. Create any necessary code/projects/assets within that directory
3. Create a bench.sh script which accepts 2 commands

• prepare to perform compilation or similar steps - this can be a nop for languages that don't need it, but it must be present
• run to actually run a benchmark - the file to run will be provided as the second argument to the script (as the first will be run)

4. Test it and open a PR

Results

Directory	Time
C++	27.7329
C	29.3115
Go	30.7541
Rust	31.0722
Java	32.0129
Swift	36.3749
C#	41.6299
JavaScript	95.9078
Ruby	4146.0210
Python	6163.0700
Bash	~150 days