vecxt/Benchmarks/logical operations

You may need to refresh the page. https://github.com/scala/scala3/issues/21637

Here is the comparison of the standard while loop with the vectorised version, when counting the true values in a boolean array.

import vecxt.plot.*
import vecxt.facades.*
showJsDocs.fromSpec(BenchmarkPlots.countTrueBenchmark, node)

Doing an "and" operation

import vecxt.plot.*
import vecxt.facades.*
showJsDocs.fromSpec(BenchmarkPlots.andBooleanBenchmark, node)

Doing or operations

import vecxt.plot.*
import vecxt.facades.*
showJsDocs.fromSpec(BenchmarkPlots.OrBooleanBenchmark, node)

Finally a double comparison, resulting in a boolean array.

import vecxt.plot.*
import vecxt.facades.*
showJsDocs.fromSpec(BenchmarkPlots.lteBenchmark, node)

Conclusion

Boolean processing gains massive benefits from SIMD, so boolean logicals like AND and OR see monster performance gains. It's a fairl niche workload though :-).

On the flip side, the comparison on two doubles, the current SIMD implementation turns out to be far slower.

In this article

Generated with