An FPGA implementation of high-throughput key-value store using Bloom filter
Jae Min Cho, Ki‐Young Choi · 2014
This paper presents an efficient implementation of key-value store using Bloom filters on FPGA. Bloom filters are used to reduce the number of unnecessary accesses to the hash tables, thereby improving the performance. Additionally, for better hash table utilization, we use a modified cuckoo hashing algorithm for the implementation. They are implemented in FPGA to further improve the performance. Experimental results show significant performance improvement over existing approaches.