FHERMA
A comprehensive platform for benchmarking FHE implementations, hosting challenges, and building the ecosystem of privacy-preserving computation.
FHERMA is the premier platform for FHE benchmarking, challenges, and ecosystem development. We provide standardized benchmarks, real-world challenges, and a community-driven approach to advancing fully homomorphic encryption technology.
The FHERMA Approach
Community-driven and challenge-based development of FHE components through structured, merit-based challenges and transparent benchmarking.
How It Works
FHERMA publishes challenges
Well-defined challenges like "efficient encrypted matrix multiplication"
Participants submit solutions
Automatically evaluated for correctness, security, and performance
Top solutions integrated
Released under Apache 2.0 license and integrated into Polycircuit
Challenge Types
Black-box Challenges
Only encrypted outputs are submitted, preserving implementation confidentiality
White-box Challenges
Full source code submitted for benchmarking and deeper analysis
Why This Model Is Effective
Efficiency
Broad range of strategies explored, often surpassing compiler-generated solutions
Openness
All contributions released under permissive licensing
Ecosystem
Welcomes both experts and new contributors
Scalability
Supports multiple FHE schemes and grows with new techniques
FHERMA Cookbook
A curated collection of reusable, high-level FHE components developed through formal challenges on the FHERMA platform.
First Edition Published
The cookbook includes components developed through formal challenges, with contributions from over 20 researchers across more than 10 leading international institutions.
Key Goals
Provide reusable and extensible component base for FHE applications
Offer detailed documentation of design strategies and trade-offs
Establish benchmarkable reference for evaluating performance and correctness
Act as foundation for standardization in component-level FHE development
Supported Libraries
Recent Challenges
Latest FHE challenges completed by the community and integrated into the ecosystem.
Recent Components
Latest FHE components developed through community challenges and integrated into the ecosystem.
String Search
Efficient algorithm for finding patterns within encrypted data while maintaining privacy guarantees.
Array Sorting
Privacy-preserving sorting algorithms that work directly on encrypted arrays without decryption.
Parity Functions
Advanced bit manipulation functions for encrypted data processing and analysis.
House Price Prediction
Predict house prices using encrypted real estate data while maintaining privacy of sensitive financial information.
Ethereum Fraud Detection
Detect fraudulent transactions on Ethereum using encrypted transaction data without revealing user privacy.
CIFAR-10 Classification
Classify images from CIFAR-10 dataset using encrypted neural networks while preserving data privacy.
What's Next
FHERMA continues to evolve as new challenges are launched and new contributors join. The platform is designed to be extensible and grow with the FHE ecosystem.
Platform Evolution
New schemes, hardware backends, and AI applications
Continuous expansion of the FHERMA Cookbook
Growing community of researchers and developers
Broader Infrastructure
In parallel, Fair Math is building a broader infrastructure layer with advanced cryptographic infrastructure and privacy-native frameworks designed specifically to empower AI and decentralized systems at scale.
"Together, these efforts will help unlock a future where privacy-preserving AI and decentralized technologies are the norm, not the exception."
Ready to Contribute?
Join the FHERMA community and start contributing to the future of privacy-preserving computation. Visit the platform to explore current challenges and get involved.