Engineering
Hardware Requirements
Minimum and recommended specifications for Sovereign Node deployment
Hardware Requirements 🖥️
To achieve "Latency Zero" and maintain high-throughput vision processing, the OBrain Sovereign Engine requires dedicated local hardware.
Recommended: Apple Silicon (M4 Series)
The engine is highly optimized for the Unified Memory Architecture and Neural Engine found in the M4 chips.
- Processor: Apple M4 or higher.
- Memory: 16GB Unified Memory (Minimum) / 32GB (Recommended for 50+ concurrent threads).
- Storage: 50GB available NVMe SSD (for image cache and local vector embeddings).
- OS: macOS 14.0+ (Sonoma) or specialized Linux kernels for ARM.
Alternative: x86_64 High-Performance
For operators using traditional server hardware.
- Processor: Intel Core i7/i9 (12th Gen+) or AMD Ryzen 7/9 with AVX-512 support.
- Memory: 32GB DDR5 RAM.
- GPU: NVIDIA RTX 3060+ (8GB VRAM) for accelerated Vision/VLM tasks (Optional but recommended).
- OS: Ubuntu 22.04 LTS.
Resource Allocation (Heuristics)
The engine dynamically scales based on available resources:
| Component | CPU Usage | RAM Usage |
|---|---|---|
| Orchestrator Thread | ~2% | < 100MB |
| Vision Worker (VLM) | ~15% per thread | ~1.2GB per model |
| Waterfall Scraper | ~5% per instance | ~400MB (Playwright Headless) |
Operational Note: Running the engine on shared Cloud VPS (vCPU) is not recommended for production due to predictable I/O throttling which breaks the "Anticipation Gap" timing.