Building Private And Secure Applications Easily Using The Sunscreen FHE Compiler&Rust
summary
Claims a usable pipeline that compiles Rust into real FHE programs using the Sunscreen FHE compiler and runs demos on AWS. Shows concrete types like Cipher<T>, integer/fractional/rational support, and schemes BFB (with limits: no encrypted comparisons, fixed-size loops, noise budget) plus work toward Taus for comparisons. Runs working examples: private ML inference (credit card fraud), blind auctions, and dark pools, with performance and correctness caveats discussed.
AI-written summary · about
chapters (15)
- 0:00Sunscreen FHE Compiler intro
- 1:05Traditional data processing risks
- 4:14What fully homomorphic encryption enables
- 6:02Traditional vs FHE model
- 7:05Why FHE isn’t widely used yet
- 8:10Sunscreen FHE compiler
- 10:57Data types in Sunscreen
- 13:50Mixing encrypted values
- 14:50Sunscreen limitations
- 16:53New comparison library
- 17:15Private ML inference
- 18:41Fraud detection demo
- 20:27Blind auctions
- 21:30Dark pools
- 25:09FHE benefits & conclusion
related talks
56mBuilding a compiler and processor for fast processing of encrypted data
39mKeynote: Towards High-Assurance Cryptographic Software
16mRust, For Science!
09mFormally Secure Hardware with Rust
20mFixing the rusty core of the internet with functional prog.
17mFHE Rollups and Co-Processor in Action
45mBuilding Safe and Secure Systems in Rust
19mProduction Unsafe Rust