threads generally slow things down. Pharo uses green threads so it does not take advantage of multiple cores. You can take advantage of multiple cores by having multiple instances of the Pharo executable running and have them communicate via sockets, pipes , shared memory etc. Some calculations / processing can also benefit from GPU acceleration , if you feel really adventurous, GPUs dont only offer concurrent coding but also speed boost compared to cpu can be anything between 8x to 100x times. Of course far less straightforward than a simple fork and it wont benefit any case scenario.��