Project Lead: Dr. Bryan Jacobs
Sponsoring Organization: DARPA
Project Synopsis: DPRIVE seeks to enable fully homomorphic encryption (FHE) computation within a factor of ten of unencrypted computations, enabling data security for all states of data – in transit, at rest, in storage, and during processing and computing. Currently, running FHE in software on standard processing hardware is infeasible for practical data security applications due to the massive processing overhead. DPRIVE aims to design and develop a hardware accelerator that integrates novel architectural approaches, hardware and software to dramatically reduce the processing overhead required to implement FHE calculations. Specifically, DPRIVE will attempt to craft new approaches to memory management; flexible data structures and programming models; and formal verification methods that ensure the design of the FHE implementation is effective and accurate, while also dramatically decreasing the overhead penalty required for FHE computations. The goal is for the accelerator to reduce the computational run time overhead by many orders of magnitude compared to current software-based FHE computations on conventional CPUs, and accelerate FHE calculations to within one order of magnitude of current performance on unencrypted data.