High-performance cryptographic random number generator (CSPRNG) implementing the AUR3ES algorithm with SHA-512 key derivation.
AUR3ES algorithm with SHA-512 key derivation, 512-bit internal state, and multiple entropy sources.
Up to 1.2 GB/s with 8-thread parallel generation. SIMD optimized for SSE2, AVX2, and AVX-512.
Works on Linux, Windows, macOS, and supports x86_64, ARM64, and ARM32 architectures.
Passed Dieharder, PractRand (64GB), TestU01 BigCrush, and NIST SP 800-22 test suites.
Secure memory zeroing, mlock() to prevent swapping, constant-time operations, entropy validation.
Built-in statistics, RNG health monitoring, performance metrics, and diagnostic tools.
Benchmark results on modern 8-core CPU @ 3.5GHz
Golden RNG has been exhaustively tested using industry-standard test suites.
| Test Suite | Tests | Result |
|---|---|---|
| Dieharder | 17 | ✅ ALL PASSED |
| PractRand | 2098+ | ✅ ALL PASSED (up to 64 GB) |
| TestU01 SmallCrush | 15 | ✅ ALL PASSED |
| TestU01 Crush | 144 | ✅ ALL PASSED |
| TestU01 BigCrush | 160 | ✅ ALL PASSED |
| NIST SP 800-22 | 15 | ✅ ALL PASSED |
| Advanced Built-in | 10 | ✅ ALL PASSED |
Total: 2,500+ statistical tests - ALL PASSED
git clone https://github.com/phenixsam/aur3es.git
cd aur3es/src
make
# Generate random bytes to stdout
./goldenrng
# Generate 10MB to file
./goldenrng -m file -o random.bin -s 10M
# Run full test suite
./goldenrng -m advanced
# Parallel benchmark
./goldenrng -m benchmark -s 100M -t 4
| Option | Description |
|---|---|
-m, --mode |
Operation mode (stdout, file, test, benchmark, simd, kat, unittests, advanced) |
-o, --output |
Output file name |
-s, --size |
Size to generate (1G, 100M, 500K) |
-t, --threads |
Number of threads for parallel benchmark |
-S, --seed |
Seed for deterministic mode |
-v, --verbose |
Enable verbose output |
-V, --version |
Show version information |
Download the source code and start generating cryptographically secure random numbers.
View on GitHub Releases