Homomorphic encryption lets cloud providers compute on encrypted data without seeing it, but the overhead is often prohibitive. A new arXiv paper introduces HULK, an algorithm designed to speed up bulk bitwise operations—AND, OR, XOR—on such encrypted data. These operations are a significant bottleneck for applications like database analytics, graph analytics, and genome analysis, which frequently perform them in bulk.
The paper's abstract notes that these applications spend considerable time on bulk bitwise operations. HULK aims to make these operations faster, potentially improving the practicality of homomorphic encryption for real-world cloud workloads. The details of the algorithm are not fully described in the abstract, but the claim is that it is efficient.
As a single source, this is a preliminary report. No independent verification or comparison with other methods is provided in the abstract.