February 2022 Summaries

The article explores the implementation of encrypted search operations using Zama’s Concrete Boolean library, enhanced by optical Fourier transform technology. It addresses the challenges of data privacy, especially in sectors like banking, where sensitive data must remain secure during processing. Traditional encryption methods require data decryption for computation, posing security risks. Fully Homomorphic Encryption (FHE) offers a solution, allowing computations on encrypted data without decryption, but is often slow. The article highlights how optical computing can accelerate FHE, enabling faster and secure data processing. Using the Concrete Boolean library, the authors demonstrate an encrypted string search process, converting character-based searches into Boolean operations to maintain privacy. While traditional search methods like tries are efficient, they are challenging to implement in FHE due to the lack of conditional branching. The study suggests that optical computing can significantly speed up FHE processes, potentially reducing computation time by a factor of 100. Through this approach, the article illustrates the potential of combining FHE with optical technology to enhance data security and processing speed, offering a glimpse into the future of encrypted data management.

Zama has released the release candidate of the 1.0.0 version of Concrete-core, a low-level crypto library designed to support fully homomorphic encryption (FHE) with an enhanced architecture for hardware acceleration. Initially focused on CPUs, the framework now allows for easier integration with various hardware accelerators, such as GPUs and FPGAs, thanks to the use of Rust's Trait system and a specification module that defines the FHE scheme. This new architecture simplifies the process for hardware developers to plug in specific implementations and verify their compliance with the FHE framework. Additionally, Zama introduces a generic benchmarking facility, Concrete-benchmark, to quickly assess the performance of new implementations, and plans to release a Concrete-test application for verifying correctness, aiming to lower the barriers for accelerating FHE and making it more mainstream. This release invites collaboration among compiler makers, library writers, and hardware designers to advance FHE technology, with Zama encouraging contributions and offering opportunities for those interested in safeguarding privacy through encryption.