3 ms·
I'm also getting a 30% speedup simply from manipulating &PyList references instead of Vecs (without parallelism).
by duckerude 5y ago
I'm also getting a 30% speedup simply from manipulating &PyList references instead of Vecs (without parallelism).