Documentation/FHE16

Developer documentation

FHE16 paper reviews, product paths, and API surface.

This page is the public entry point for FHE16 technical documentation. The docs are split into paper reviews, product applications, and the FHE16 API so each area can grow independently.

A. PapersReview the FHE16 research baseBootstrapping, arithmetic, correctness, and MPC context.B. ProductsMap research to product workflowsBlockchain confidential state, auctions, and private scoring.C. APIDefine one method surfaceC ABI first, then Web Browser, Rust, and Python bindings.

Documentation model

The docs should read like technical product infrastructure, not a publication list. Each page ties research evidence to the methods, parameters, and trust boundaries a developer needs.

ConceptEngine

FHE16 as the engine layer

Explain the role of encrypted integer operations, bootstrapping, parameters, and benchmark profiles before exposing product APIs.

ContractFFI

One API contract across languages

Keep method semantics stable across C, Web Browser/WASM, Rust, and Python. Language wrappers can be idiomatic without changing behavior.

A. FHE16 paper reviews

Each paper gets a structured review: what the paper proves, what it means for the engine, what API behavior it supports, and what still needs implementation evidence.

FHE16Bootstrapping

Fast, Compact and Hardware-Friendly Bootstrapping

Anchor paper for the low-latency FHE16 narrative and the benchmark claim that encrypted computation can be practical for product workloads.

ArithmeticMethods

Parallel prefix FHE arithmetic

Connects arithmetic circuits to API methods such as add, lt, max, and select.

CorrectnessParameters

Blind rotation error analysis

Supports correctness, noise budget, and parameter documentation for reliable FHE16 execution profiles.

MPCThreshold

Actively secure MPC

Frames SSFHE, threshold disclosure, and multi-party workflows beyond a single-party encrypted compute setup.

B. Products built with FHE16

Product docs should show which encrypted operations are used, what data stays private, and where a user or verifier sees the result.

BlockchainConfidential state

Encrypted state and selective disclosure

Compute or verify over encrypted state while allowing threshold or policy-based disclosure when a protocol requires a result.

AuctionComparison

Confidential auction workflows

Keep bids hidden while computing winner selection with encrypted comparison, max, and conditional selection methods.

EnterpriseRisk scoring

Private risk scoring

Score sensitive insurance, identity, account, or transaction data without exposing raw records to the compute operator.

B2CDemand

Private demand intelligence

Aggregate or compare encrypted preference signals while limiting raw customer data movement across the product stack.

C. FHE16 API

The API docs now start from FHE16Ver6's current C ABI. The API overview stays short, and each operation has its own page with caller, circuit, and ownership notes.

ABIFHE16Ver6 / soAPI.hpp

Draft OpenAPI included

Core methods stay consistent across language bindings. Each method also needs a caller definition and a circuit path so product teams know who invokes it and what homomorphic structure runs.

FHE16_GenEvalruntime and key init
FHE16_ENCIntinteger encryption
FHE16_DECIntinteger decryption
ADD / SUB / SMULLarithmetic
LT / LE / GT / GE / EQcomparison
MAX / MIN / SELECTcompose

Open the FHE16 API page · Browse operation pages

// Current C ABI direction.
int32_t* sk = FHE16_GenEval();
int32_t* a = FHE16_ENCInt(42, 8);
int32_t* b = FHE16_ENCInt(100, 8);

int32_t* less = FHE16_LT(a, b);
int64_t opened = FHE16_DECInt(less, sk);

FHE16_Free(less);

Language bindings

The docs source separates binding pages so each SDK can grow with its own examples, ownership rules, and API notes.

BrowserWASM

Web Browser

Client-side demos and browser-native encrypted computation experiments using a WASM wrapper.

CABI

C

The lowest stable boundary for FFI, integration tests, and wrappers in other languages.

RustSafe wrapper

Rust

Typed resource management over the C ABI with explicit ownership and error handling.

PythonPrototype

Python

Research notebooks, product prototypes, and benchmark orchestration against the same method semantics.