MetaMUI Crypto Primitives
Pure post-quantum cryptography with hardware acceleration. No dependencies.
48
Algorithms
10
Platforms
6
SIMD Backends
NIST + KPQC
Standards
Quick Start
git clone https://github.com/SovereignWallet-Network/metamui-crypto-primitives.git
cd metamui-crypto-primitives
# Python
cd metamui-crypto-python && python3 -m pytest tests/ -v
# Rust
cd metamui-crypto-rust && cargo build --release
# TypeScript
cd metamui-crypto-typescript && npm install && npm run build
# Swift
cd metamui-crypto-swift && swift build
# Kotlin
cd metamui-crypto-kotlin && ./gradlew build
# Go
cd metamui-crypto-go && go build ./...
# C
cd metamui-crypto-c && cmake -B build && cmake --build build
# C# / Java
cd metamui-crypto-csharp && dotnet build
cd metamui-crypto-java && mvn compile
Hardware Acceleration
Automatic runtime detection selects the fastest available backend per platform. Zero configuration required.
| Backend | Architecture | Algorithms | Register Width |
|---|---|---|---|
| AVX-512 | Intel Ice Lake+, AMD Zen 4+ | Falcon NTT/FFT, BLAKE3 batch | 512-bit (32 x u16) |
| AVX-2 | x86-64 (Haswell 2013+) | Falcon, BLAKE3, ML-KEM, HAETAE | 256-bit (16 x u16) |
| ARM NEON | AArch64 (all ARM64) | Falcon, BLAKE3, ML-KEM | 128-bit (8 x u16) |
| Apple Metal | Apple Silicon M1-M4 | Falcon NTT, BLAKE3, SMAUG-T | GPU compute shaders |
| NVIDIA CUDA | CC 8.9+ (Ada Lovelace) | BLAKE3 batch, SMAUG-T | GPU parallel kernels |
| Portable | Any platform | All 48 algorithms | Scalar fallback |
Dispatch hierarchy: Metal GPU → AVX-512 → AVX-2 → NEON → Portable
All SIMD paths are constant-time with branchless masking. Learn more →
Getting Started
Build from source for your platform
Hardware Acceleration
AVX-512, NEON, Metal GPU, CUDA
Algorithms
48 algorithms across 12 categories
API Reference
Documentation for all 10 platforms
Security
Constant-time, side-channel resistance
Interoperability
Cross-language verification across 10 platforms
Security
- Constant-time implementations - All operations use branchless, constant-time code paths
- Secure memory clearing - Automatic zeroization of sensitive key material
- Side-channel resistance - Protection against timing, cache, and power analysis attacks
- IND-CCA2 security - Fujisaki-Okamoto transform for all KEM constructions
- Test vector validation - Verified against NIST CAVP/ACVP and KPQC KAT vectors
Supported Platforms
| Platform | Module | SIMD |
|---|---|---|
| Rust | metamui-crypto-rust |
AVX-512, AVX-2, NEON, Metal, CUDA |
| C | metamui-crypto-c |
AVX-512, AVX-2, NEON, Metal, CUDA |
| Swift | metamui-crypto-swift |
NEON, Metal |
| Go | metamui-crypto-go |
Metal, CUDA |
| Python | metamui-crypto-python |
Metal, CUDA (via Cython) |
| TypeScript | metamui-crypto-typescript |
WASM SIMD128 |
| Kotlin | metamui-crypto-kotlin |
JVM Vector API |
| Java | metamui-crypto-java |
JVM Vector API |
| C# | metamui-crypto-csharp |
ARM NEON intrinsics |
| WASM | metamui-crypto-wasm |
SIMD128 |
Algorithms
Post-Quantum Cryptography (19)
NIST Standardized — ML-KEM-512/768/1024 (FIPS 203), ML-DSA-44/65/87 (FIPS 204), SLH-DSA-SHA2-128f / SLH-DSA-SHAKE-256f (FIPS 205), Falcon-512/1024
Korean PQC (KPQC) — SMAUG-T (T1/T3/T5), HAETAE (2/3/5), AIMer (128f/192f/256f), NTRU+ (576/768/864/1152)
Additional PQC — NTRU Prime, FrodoKEM-640/976, HQC-128, Classic-McEliece, XMSS, LMS
Classical Cryptography (29)
Digital Signatures — Ed25519, Ed25519-ZIP215, Sr25519, RSA-2048
Symmetric & AEAD — AES-256-GCM, AES-256-CTR, ChaCha20, ChaCha20-Poly1305, ARIA-256, Camellia-256, Ascon-128/128a
Hash Functions — SHA-256/384/512, SHA3-256/512, BLAKE2b-256/512, BLAKE3, Keccak-256, SipHash, FlatHash
Key Derivation — Argon2id, PBKDF2, HKDF-SHA256, BLAKE3-KDF
MAC & Utilities — HMAC-SHA256/512, Poly1305, CMAC, X25519, HMAC-DRBG, BIP39
Licensed under BSL 1.1 / Commercial - Sovereign Wallet Co., Ltd.