Adapting the Bloom filter to multithreaded environments
Ivan Voras, Mario Žagar · 2010
Many classical algorithms like the Bloom filter were envisioned and created at a time when computers were the size of rooms and multithreading was not yet even explored theoretically. The landscape of modern mainstream computer systems today is heavily dominated with multi-core CPU-s but the effort to make full use of such systems is still ongoing. The Bloom filter has proven itself useful both as a core algorithm and as a supportive or optimizing addition to other data classification algorithms. This work explores adaptations to the basic Bloom filter algorithm for use in multithreaded applications on contemporary SMP systems and the consequences of such adaptations to its overall efficiency.