AI in the Black Sea
AI in the Black Sea: Server Configuration Documentation
This document details the server configuration powering the "AI in the Black Sea" project. This project utilizes a distributed server network to process and analyze real-time data feeds related to maritime activity, environmental conditions, and potential security threats within the Black Sea region. This guide is intended for new system administrators and developers joining the project. Understanding this configuration is crucial for maintenance, scaling, and troubleshooting.
Overview
The "AI in the Black Sea" system comprises three distinct tiers: Data Acquisition, Processing, and Analytical. Each tier is housed on dedicated server clusters, optimized for their respective tasks. The project leverages a hybrid cloud approach, utilizing both on-premise hardware for low-latency data ingestion and cloud-based resources for scalable processing and storage. Network connectivity is paramount, relying on redundant fiber optic links and robust security protocols detailed in the Security Protocols documentation. The core AI models are detailed in the AI Model Documentation section. Initial deployment was completed on 2024-02-29. See Project History for a timeline.
Data Acquisition Tier
This tier is responsible for ingesting data from a variety of sources, including: Automatic Identification System (AIS) transponders, satellite imagery, weather sensors, and open-source intelligence (OSINT) feeds. Low latency is critical here. Servers in this tier are geographically distributed around the Black Sea coastline for optimal coverage. See Data Sources for complete list.
Hardware Specifications
| Component | Specification | |
|---|---|---|
| CPU | Intel Xeon Gold 6338 (32 cores/64 threads) | |
| RAM | 256 GB DDR4 ECC Registered (3200 MHz) | |
| Storage | 2 x 1TB NVMe SSD (RAID 1) – OS & Application Data | 4 x 8TB SAS HDD (RAID 10) – Raw Data Storage |
| Network Interface | Dual 100GbE NICs (Redundant) | |
| Power Supply | 2 x 1600W Redundant Power Supplies |
Software Stack
- Operating System: CentOS Stream 9
- Database: PostgreSQL 14 with PostGIS extension for geospatial data. See Database Schema
- Message Queue: RabbitMQ 3.9 for asynchronous data ingestion.
- Data Ingestion Software: Custom Python scripts leveraging libraries like `pcapy` for AIS decoding and `GDAL` for satellite image processing. See Ingestion Scripts
- Monitoring: Prometheus and Grafana for system health monitoring.
Processing Tier
This tier performs the initial processing and filtering of the ingested data. This includes data cleaning, normalization, and feature extraction. The majority of this tier resides in a private cloud environment managed with OpenStack.
Hardware Specifications
| Component | Specification | |
|---|---|---|
| CPU | AMD EPYC 7763 (64 cores/128 threads) | |
| RAM | 512 GB DDR4 ECC Registered (3200 MHz) | |
| Storage | 1 x 2TB NVMe SSD – OS & Application Data | 8 x 16TB SAS HDD (RAID 6) – Processed Data Storage |
| GPU | 4 x NVIDIA A100 (40GB) – Accelerated Processing | |
| Network Interface | Dual 40GbE NICs |
Software Stack
- Operating System: Ubuntu Server 22.04
- Data Processing Framework: Apache Spark 3.3 with PySpark for distributed data processing. See Spark Configuration
- Machine Learning Libraries: TensorFlow 2.10, PyTorch 1.13
- Containerization: Docker and Kubernetes for application deployment and scaling. See Kubernetes Manifests
- Job Scheduling: Airflow for workflow management.
Analytical Tier
This tier focuses on running the AI models to generate insights and alerts. This tier is highly scalable, leveraging a public cloud provider (currently Amazon Web Services). Results are visualized on a dedicated Dashboard.
Hardware Specifications (Cloud Instances)
| Instance Type | vCPUs | Memory (GB) | GPU |
|---|---|---|---|
| AWS p4d.24xlarge | 96 | 1152 | 8 x NVIDIA A100 (40GB) |
| AWS r6a.48xlarge | 192 | 768 | None |
Software Stack
- Operating System: Amazon Linux 2
- Machine Learning Framework: TensorFlow 2.10 and PyTorch 1.13 (same as Processing Tier)
- Data Storage: Amazon S3 for long-term data archiving.
- API Gateway: Amazon API Gateway for providing access to AI model outputs.
- Monitoring: Integrated with AWS CloudWatch.
Inter-Tier Communication
Communication between tiers is secured using TLS/SSL encryption. Data transmission utilizes a combination of REST APIs and message queues. Detailed API specifications can be found in the API Documentation.
Future Expansion
Planned expansions include integrating additional data sources (e.g., sonar data) and upgrading to newer hardware generations. See Roadmap for planned changes.
Intel-Based Server Configurations
| Configuration | Specifications | Benchmark |
|---|---|---|
| Core i7-6700K/7700 Server | 64 GB DDR4, NVMe SSD 2 x 512 GB | CPU Benchmark: 8046 |
| Core i7-8700 Server | 64 GB DDR4, NVMe SSD 2x1 TB | CPU Benchmark: 13124 |
| Core i9-9900K Server | 128 GB DDR4, NVMe SSD 2 x 1 TB | CPU Benchmark: 49969 |
| Core i9-13900 Server (64GB) | 64 GB RAM, 2x2 TB NVMe SSD | |
| Core i9-13900 Server (128GB) | 128 GB RAM, 2x2 TB NVMe SSD | |
| Core i5-13500 Server (64GB) | 64 GB RAM, 2x500 GB NVMe SSD | |
| Core i5-13500 Server (128GB) | 128 GB RAM, 2x500 GB NVMe SSD | |
| Core i5-13500 Workstation | 64 GB DDR5 RAM, 2 NVMe SSD, NVIDIA RTX 4000 |
AMD-Based Server Configurations
| Configuration | Specifications | Benchmark |
|---|---|---|
| Ryzen 5 3600 Server | 64 GB RAM, 2x480 GB NVMe | CPU Benchmark: 17849 |
| Ryzen 7 7700 Server | 64 GB DDR5 RAM, 2x1 TB NVMe | CPU Benchmark: 35224 |
| Ryzen 9 5950X Server | 128 GB RAM, 2x4 TB NVMe | CPU Benchmark: 46045 |
| Ryzen 9 7950X Server | 128 GB DDR5 ECC, 2x2 TB NVMe | CPU Benchmark: 63561 |
| EPYC 7502P Server (128GB/1TB) | 128 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (128GB/2TB) | 128 GB RAM, 2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (128GB/4TB) | 128 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (256GB/1TB) | 256 GB RAM, 1 TB NVMe | CPU Benchmark: 48021 |
| EPYC 7502P Server (256GB/4TB) | 256 GB RAM, 2x2 TB NVMe | CPU Benchmark: 48021 |
| EPYC 9454P Server | 256 GB RAM, 2x2 TB NVMe |
Order Your Dedicated Server
Configure and order your ideal server configuration
Need Assistance?
- Telegram: @powervps Servers at a discounted price
⚠️ *Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock.* ⚠️