Getting Started - NEUROVEDIK

📋 Requirements

Before installing NEUROVEDIK, ensure you have the following:

Operating System

Windows 10/11 (x64)
Linux (Ubuntu 20.04+, CentOS 8+)
macOS 12+ (Intel only)

Build Tools

CMake 3.16+
GCC 10+ / Clang 12+ / MSVC 2019+
Python 3.7+ (for bindings)

Recommended Hardware

Intel Skylake-X or newer (AVX-512)
AMD Zen 4 (AVX-512 support)
8GB RAM minimum

💡

Note: NEUROVEDIK includes scalar fallbacks, so it will work on any x64 CPU. AVX-512 is optional but provides the best performance.

📦 Installation

Option 1: Clone from GitHub

bash

# Clone the repository
git clone https://github.com/neurovedik/neurovedik.git
cd neurovedik

# Verify the structure
ls -la

Option 2: Download Release

Download the latest release from GitHub Releases.

🔨 Building from Source

cmd

REM Run the build script
build.bat

REM Or use CMake directly
mkdir build
cd build
cmake .. -G "Visual Studio 17 2022" -A x64
cmake --build . --config Release

bash

# Install dependencies
sudo apt install cmake gcc g++ python3-dev

# Build
mkdir build && cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make -j$(nproc)

# Install (optional)
sudo make install

bash

# Install Xcode command line tools
xcode-select --install

# Install CMake via Homebrew
brew install cmake

# Build
mkdir build && cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
make -j$(sysctl -n hw.ncpu)

🐍 Python Setup

Install the Python bindings for easy integration with your data science workflow:

bash

# Install from source (in the project root)
pip install -e .

# Verify installation
python -c "import NEUROVEDIK; print(NEUROVEDIK.get_cpu_features())"

✅

If you see CPU feature information, the installation was successful!

🚀 Your First Program

Python Example

python hello_neurovedik.py

import NEUROVEDIK

# Create large integers
a = NEUROVEDIK.BigInt.from_int(123456789012345678901234567890)
b = NEUROVEDIK.BigInt.from_int(987654321098765432109876543210)

# Multiply using Vedic algorithms
result = NEUROVEDIK.multiply(a, b)
print(f"Result: {result.to_int()}")

# Check which SIMD features are available
features = NEUROVEDIK.get_cpu_features()
print(f"AVX-512: {features['avx512f']}")
print(f"AVX2: {features['avx2']}")

# RSA-style modular exponentiation
base = NEUROVEDIK.BigInt.from_int(12345)
exp = NEUROVEDIK.BigInt.from_int(65537)
mod = NEUROVEDIK.BigInt.from_int(1000000007)

cipher = NEUROVEDIK.modexp(base, exp, mod)
print(f"ModExp result: {cipher.to_int()}")

C Example

c hello_neurovedik.c

#include "neurovedik.h"
#include <stdio.h>

int main() {
    // Allocate BigInts
    BigInt* a = nv_alloc(32);  // 32 limbs = 2048 bits
    BigInt* b = nv_alloc(32);
    BigInt* result = nv_alloc(64);
    
    // Initialize with values
    nv_from_hex(a, "123456789ABCDEF...");
    nv_from_hex(b, "FEDCBA987654321...");
    
    // Multiply (auto-selects best algorithm)
    nv_multiply(a, b, result);
    
    // Print result
    char* hex = nv_to_hex(result);
    printf("Result: %s\n", hex);
    
    // Cleanup
    nv_free(a);
    nv_free(b);
    nv_free(result);
    
    return 0;
}

✅ Verification

Run the verification suite to ensure everything works correctly:

bash

# Run Python verification
python verify_algorithms.py

# Expected output:
# ✓ All 56 tests passed!
# ✓ CPU Features: AVX-512 detected
# ✓ Algorithms verified: Vedic, Karatsuba, NTT, Montgomery