AI in the Ural Mountains
AI in the Ural Mountains: Server Configuration
This article details the server configuration for our "AI in the Ural Mountains" project, a distributed computing initiative focused on processing geological data using machine learning algorithms. This guide is aimed at new team members responsible for server maintenance and deployment. It covers hardware, software, networking, and security aspects of the system.
Overview
The project utilizes a cluster of servers located in a secure facility within the Ural Mountains. The primary goal is to analyze seismic data, mineral composition scans, and historical geological surveys to identify potential resource deposits and predict geological events. The server architecture is designed for high throughput, scalability, and redundancy. The operating system of choice is Ubuntu Server 22.04 LTS, due to its stability, community support, and compatibility with the necessary machine learning frameworks.
Hardware Configuration
The cluster consists of 20 identical servers, with one designated as the master node. Each server is built with the following specifications:
| Component | Specification |
|---|---|
| CPU | AMD EPYC 7763 (64 Cores, 128 Threads) |
| RAM | 256 GB DDR4 ECC Registered RAM |
| Storage (OS) | 1 TB NVMe SSD |
| Storage (Data) | 16 TB SAS HDD (RAID 6) |
| Network Interface | Dual 100 GbE Ethernet |
| Power Supply | Redundant 1600W Platinum PSUs |
The master node has slightly enhanced specifications for coordinating the cluster. These are detailed below:
| Component | Specification |
|---|---|
| CPU | AMD EPYC 7763 (64 Cores, 128 Threads) |
| RAM | 512 GB DDR4 ECC Registered RAM |
| Storage (OS) | 2 TB NVMe SSD (RAID 1) |
| Storage (Data) | 32 TB SAS HDD (RAID 6) |
| Network Interface | Quad 100 GbE Ethernet |
A dedicated Network Attached Storage (NAS) device with 1PB of capacity is used for long-term data archiving. All servers are housed in a temperature and humidity-controlled data center with redundant power and cooling systems. See Data Center Redundancy for more details.
Software Stack
Each server runs a standardized software stack, ensuring consistency and ease of management.
| Software | Version | Purpose |
|---|---|---|
| Operating System | Ubuntu Server 22.04 LTS | Base OS |
| Python | 3.10 | Primary programming language |
| TensorFlow | 2.12 | Machine Learning Framework |
| PyTorch | 2.0 | Alternative Machine Learning Framework |
| CUDA Toolkit | 12.1 | GPU Acceleration |
| Docker | 20.10 | Containerization |
| Kubernetes | 1.26 | Container Orchestration |
| SSH Server | OpenSSH 8.2 | Remote Access |
We utilize Docker and Kubernetes for containerization and orchestration, allowing for efficient resource utilization and simplified deployment of machine learning models. The master node also runs a Prometheus instance for monitoring and alerting. Detailed instructions for setting up the software stack are available on the Software Installation Guide page.
Networking Configuration
The servers are connected via a dedicated 100 GbE network. The network topology is a Clos network, providing high bandwidth and low latency. A dedicated VLAN is used for inter-server communication, and another for external access. The master node acts as the network gateway. Firewall rules are configured using iptables to restrict access to essential services. The network configuration details are documented in the Network Diagram. We also employ DNS for service discovery within the cluster.
Security Considerations
Security is paramount. The following measures are in place:
- **Physical Security:** The data center is physically secured with multiple layers of access control, including biometric scanners and surveillance cameras. See Physical Security Protocols for details.
- **Network Security:** Firewalls, intrusion detection systems, and regular security audits are implemented to protect the network. Access to the network is restricted to authorized personnel.
- **Data Encryption:** All sensitive data is encrypted at rest and in transit. We use TLS/SSL for secure communication.
- **User Authentication:** Strong passwords and multi-factor authentication are required for all user accounts.
- **Regular Backups:** Regular backups of all critical data are performed and stored offsite. See the Backup and Recovery Plan for specifics.
- **Vulnerability Scanning**: Regular vulnerability scans using tools like Nessus are performed to identify and remediate security weaknesses.
Data Flow
Raw data is ingested from various sources, including seismic sensors and geological survey databases. This data is initially stored on the NAS device. The master node then distributes tasks to the worker nodes via Kubernetes. Each worker node processes a portion of the data using the designated machine learning algorithms. Results are aggregated on the master node and stored in a centralized database. See the Data Pipeline Diagram for a visual representation of the data flow.
Future Enhancements
We are planning to integrate GPU acceleration to further enhance the performance of our machine learning models. We are also exploring the use of a distributed file system like Hadoop Distributed File System (HDFS) to improve data access speeds.
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.* ⚠️