Performance

String manipulation and math operations are the heart of almost every programming task. And the core purpose of programming languages like Lasso is to perform these processes as quickly as possible. Thanks in part to its new LLVM core and native compilation, common speed tests show how Lasso 9 outperforms the competition.

How The Tests Were Performed


The numbers shown here represent the average result for each test over 100 iterations. Results are shown in seconds. The tests were performed on an Apple MacBook running MacOS X 10.5, comparing Lasso 9.0.0 against the version of each competing language that ships with the OS. Specifically:


• Python 2.5

• PHP 5.3.0

• Ruby 1.8.7

Substring Test


The substring test measures the performance of basic string manipulation by timing how long it takes to calculate every possible substring containing the first character from a string that is 130,000 characters in length. In this test, Lasso 9 demonstrates a nearly 40% improvement over the closest competitor.
 

Fractal Test


The fractal test measures the performance of floating point math operations by timing how long it takes to calculate the first 1,000 iterations of the Mandelbrot set. Again, Lasso 9 shows significant savings — a little over 20% — versus the closest competitor.
 

Fibonacci Test


The fibonacci test measures the performance of basic math operations by timing how long it takes to calculate the 40th number in the Fibonacci sequence. Here, Lasso 9 just barely edges out Python by about a half second, though both are about 15% faster than PHP and nearly 60% faster than Ruby.
 

Count Test


The count test measures the performance of looping structures by timing how long it takes to count from zero to one million. Here, somewhat surprisingly when compared to the other tests, PHP wins by a huge margin, though Lasso 9 still outperforms Python and Ruby.