AI in the Sea of Okhotsk
AI in the Sea of Okhotsk: Server Configuration
This document details the server configuration supporting the "AI in the Sea of Okhotsk" project, a long-term initiative utilizing artificial intelligence to monitor and analyze marine life and environmental conditions in the Sea of Okhotsk. This guide is intended for new system administrators and engineers joining the project, detailing the hardware, software, and network setup. It assumes familiarity with basic Linux server administration and MediaWiki syntax.
Project Overview
The "AI in the Sea of Okhotsk" project relies on a distributed network of underwater sensors and surface buoys collecting data – acoustic recordings, temperature readings, salinity measurements, and video feeds. This data is transmitted to our central server cluster for processing by a suite of AI algorithms, focused on species identification, anomaly detection, and predictive modeling. Data processing pipelines are crucial to the success of this project.
Hardware Configuration
The core server infrastructure consists of three primary nodes: a Data Acquisition Node, a Processing Node, and a Storage Node. Each node is housed in a secure, climate-controlled data center in Magadan, Russia.
| Node Type | CPU | RAM | Storage | Network Interface |
|---|---|---|---|---|
| Intel Xeon Gold 6248R (24 cores) | 128 GB DDR4 ECC | 2 x 4TB NVMe SSD (RAID 1) | 10 Gbps Ethernet | | ||||
| 2 x AMD EPYC 7763 (64 cores each) | 256 GB DDR4 ECC | 1 x 1TB NVMe SSD (OS), 4 x 8TB SATA HDD (Data Cache) | 25 Gbps Ethernet | | ||||
| Supermicro SuperStorage Server | Intel Xeon Silver 4210 (10 cores) | 32 x 16TB SATA HDD (RAID 6) | 40 Gbps InfiniBand | |
Each node runs a dedicated power distribution unit (PDU) with redundant power supplies. Uninterruptible Power Supplies (UPS) provide backup power for up to 30 minutes in case of a power outage. Remote monitoring is handled via IPMI.
Software Stack
All servers utilize a base installation of Ubuntu Server 22.04 LTS. We employ a containerized approach using Docker and Kubernetes for application deployment and management.
| Component | Version | Description |
|---|---|---|
| Ubuntu Server 22.04 LTS | Provides the base operating environment. | | ||
| Docker 24.0.5 | Enables application packaging and isolation. | | ||
| Kubernetes 1.27 | Manages container deployment and scaling. | | ||
| PostgreSQL 15 | Stores metadata and processed data. | | ||
| TensorFlow 2.13 | The primary framework for developing and deploying AI models. | | ||
| Prometheus & Grafana | Provides real-time system and application monitoring. | |
The AI models themselves are developed in Python using TensorFlow and are regularly updated through a continuous integration/continuous deployment (CI/CD) pipeline using GitLab. Security patching is performed weekly.
Network Configuration
The server cluster is connected to the internet via a dedicated 1 Gbps fiber optic connection. Internal network communication relies on a private network utilizing the 10.0.0.0/16 subnet. Firewall rules are managed using iptables and are regularly reviewed and updated to ensure security. A dedicated Virtual Private Network (VPN) allows secure remote access for administrators.
| Network Interface | IP Address | Subnet Mask | Gateway |
|---|---|---|---|
| 192.0.2.100 | 255.255.255.0 | 192.0.2.1 | | |||
| 10.0.1.10 | 255.255.255.0 | 10.0.1.1 | | |||
| 192.168.1.10 | 255.255.255.0 | 192.168.1.1 | |
Network monitoring is performed using Nagios. Regular network penetration testing is conducted to identify and address potential vulnerabilities. DNS resolution is handled by a local DNS server.
Data Flow
1. Data is collected by underwater sensors and surface buoys. 2. Data is transmitted to the Data Acquisition Node via satellite communication. 3. The Data Acquisition Node validates and pre-processes the data. 4. Pre-processed data is sent to the Processing Node for AI analysis. 5. The Processing Node runs AI models to identify species, detect anomalies, and generate predictions. 6. Processed data and results are stored in the Storage Node. 7. Researchers access the data via a web-based interface powered by Flask.
Future Considerations
Future upgrades include expanding the storage capacity of the Storage Node, upgrading the network infrastructure to 100 Gbps, and implementing a more sophisticated AI model deployment strategy using machine learning operations (MLOps). We also plan to integrate with satellite imagery analysis tools.
Server maintenance procedures are documented separately.
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.* ⚠️