WebAssemblyVSJavaScript

Project to compare the performance of WebAssembly with JavaScript.

We use some good programs in C and JavaScript written for the Computer Language Benchmarks Game.

We have kept the following benchmarks:

• binarytrees
• fannkuchredux
• fasta
• mandelbrot
• nbody
• reverse-complement (can be run on data/knucleotide-input10000.txt)
• spectralnorm

All benchmarks are single-threaded.

Method

• Compile the C code to WebAssembly. Different flags and compilers are possible.
• Execute both the JavaScript and the WebAssembly.
• Compare the performance.

Ideally memory usage, file sizes and different environments should be included in the benchmark. Most browsers support WebAssembly.

Outstanding problems

• The fasta test is problematic since it implies outputting a large volume of data to the standard output. It is likely to become a limiting factor. In JavaScript, we can use a buffered output. For the C version, we need some fast print-out approach.

References

• Egorov, Maybe you don't need Rust and WASM to speed up your JS
• Haas et al., Bringing the Web up to Speed with WebAssembly, PLDI'17
• Herrera et al., WebAssembly and JavaScript Challenge: Numerical program performance using modern browser technologies and devices, 2018
• Lemire, Is WebAssembly faster than JavaScript?