Binary Fuse Filters: Fast and Smaller Than Xor Filters

TL;DR

Binary fuse filters are a new probabilistic data structure that offers faster construction, smaller size, and comparable query speed to xor filters, making them highly efficient for approximate set membership testing.

Contribution

We introduce binary fuse filters, a novel probabilistic filter that improves construction speed and reduces storage size while maintaining fast query performance.

Findings

Binary fuse filters are within 13% of the theoretical storage lower bound.

Construction of binary fuse filters can be more than twice as fast as xor filters.

Binary fuse filters outperform xor filters and other alternatives in experiments.

Abstract

Bloom and cuckoo filters provide fast approximate set membership while using little memory. Engineers use them to avoid expensive disk and network accesses. The recently introduced xor filters can be faster and smaller than Bloom and cuckoo filters. The xor filters are within 23% of the theoretical lower bound in storage as opposed to 44% for Bloom filters. Inspired by Dietzfelbinger and Walzer, we build probabilistic filters -- called binary fuse filters -- that are within 13% of the storage lower bound -- without sacrificing query speed. As an additional benefit, the construction of the new binary fuse filters can be more than twice as fast as the construction of xor filters. By slightly sacrificing query speed, we further reduce storage to within 8% of the lower bound. We compare the performance against a wide range of competitive alternatives such as Bloom filters, blocked Bloom filters, vector quotient filters, cuckoo filters, and the recent ribbon filters. Our experiments suggest that binary fuse filters are superior to xor filters.