This month, our team at Zama released some major updates to four of our libraries. With these releases, Zama continues to build its suite of products to make fully homomorphic encryption accessible, easy, and fast.
Concrete Core v1.0
This version of Concrete-core v1.0 provides the open source community a platform to integrate FHE primitives implemented on any hardware. The aim of this platform is to experiment with hardware acceleration as swiftly as possible, to help resolve one of the main challenges facing FHE as of today: performance.
This version of Concrete is based on the recent release of Concrete-core v1.0. The concrete-core-experimental crate is now deprecated, the Fast Fourier transforms computations now relies on FFT instead of FFTW, the key generation is now parallelised, and we added native support of Apple Silicon architectures.
Concrete Numpy v0.8
This version of Concrete Numpy has been completely re-written from scratch (from Concrete Numpy v0.6) in order to reduce technical debts. It also implements new features like Python 3.7 and 3.10 Support, Large Bit Width Support on Circuits without Table Lookups, Assignment to Tensors and many more.
Concrete ML v0.4
This version of Concrete-ML adds more support for quantization aware training neural networks, adds decision tree-ensemble regressors, and includes additional linear regression models. For custom models, first-class support for Brevitas, a quantization aware training framework was added: a dedicated function imports models containing Brevitas layers directly.