| We investigate the functional programming language Haskell and its support for pure parallel programming. In this technique, the programmer adds annotations to code indicating what parts might be profitable to execute in parallel. Such annotations can never change the result of the computation, only its performance. We implemented a sort algorithm in Haskell and applied different parameters to it, and measured the performance results on a multi-core processor. We analyze the results in terms of Amdahl's Law. |
