AI in Montserrat
AI in Montserrat: Server Configuration
This article details the server configuration supporting Artificial Intelligence (AI) initiatives within Montserrat. This configuration is designed for both research and deployment of AI models, specifically focusing on applications related to environmental monitoring, disaster prediction, and educational resources. This guide is intended for newcomers to the Montserrat server infrastructure. It assumes a basic understanding of server administration and Linux operating systems.
Overview
The AI infrastructure in Montserrat relies on a clustered server environment to provide high availability, scalability, and computational power. The core of this infrastructure consists of three primary servers: 'Phoenix', 'Ember', and 'Nova'. Each server is equipped with specialized hardware to accelerate AI workloads. The entire system is networked using a dedicated 10Gbps fiber optic backbone, connected to the main Montserrat Data Network. All servers run Ubuntu Server 22.04 LTS and are managed via Ansible. Regular server backups are performed and stored off-site. Access is controlled via SSH key authentication and monitored by the Security Team.
Hardware Specifications
The following tables detail the hardware specifications for each of the primary AI servers. These specifications are critical for understanding the capabilities and limitations of the system.
| Server Name | CPU | RAM | GPU | Storage |
|---|---|---|---|---|
| Phoenix | AMD EPYC 7763 (64 cores) | 512 GB DDR4 ECC | NVIDIA RTX A6000 (48 GB) | 2 x 4TB NVMe SSD (RAID 1) |
| Ember | Intel Xeon Platinum 8380 (40 cores) | 256 GB DDR4 ECC | NVIDIA RTX A5000 (24 GB) | 1 x 8TB NVMe SSD |
| Nova | AMD EPYC 7713 (64 cores) | 256 GB DDR4 ECC | NVIDIA Tesla T4 (16 GB) | 4 x 2TB HDD (RAID 5) |
This table outlines the core hardware. Additional hardware, such as network interface cards (NICs) and power supplies, are redundant for increased reliability. Detailed hardware inventory is available on the Asset Management System.
Software Stack
The software stack is designed to support a wide range of AI frameworks and tools. All servers utilise Docker for containerization, enabling consistent deployment and management of applications. The core software components are listed below.
| Component | Version | Purpose |
|---|---|---|
| Python | 3.10 | Primary programming language for AI development. |
| TensorFlow | 2.12 | Machine learning framework. |
| PyTorch | 2.0 | Machine learning framework. |
| CUDA Toolkit | 12.2 | NVIDIA's parallel computing platform and API. |
| Jupyter Notebook | 6.4 | Interactive computing environment. |
| Docker | 20.10 | Containerization platform. |
| Docker Compose | 2.17 | Tool for defining and running multi-container Docker applications. |
The software stack is regularly updated via automated patching to ensure security and stability. A detailed software inventory is available on the Software Repository. Access to the software environment is managed via user accounts and group permissions.
Networking & Security
The servers are configured within a dedicated Virtual Local Area Network (VLAN) to isolate AI workloads from other network traffic. A firewall, configured using iptables, restricts access to essential ports only. All network traffic is monitored by the Network Monitoring System. Server access is strictly controlled through SSH key authentication and multi-factor authentication (MFA). Regular security audits are conducted to identify and address potential vulnerabilities.
| Feature | Configuration | Details |
|---|---|---|
| VLAN | 192.168.10.0/24 | Dedicated network for AI servers. |
| Firewall | iptables | Restricts inbound and outbound traffic. |
| SSH Access | Key-based authentication | Password authentication is disabled. |
| MFA | Google Authenticator | Required for all administrative access. |
| Intrusion Detection | Suricata | Monitors network traffic for malicious activity. |
This setup prioritizes security and network segregation. Further information on network and security protocols is available in the Security Documentation.
Future Expansion
Future expansion plans include adding additional GPU servers and increasing storage capacity. We are also exploring the implementation of a distributed computing framework, such as Dask, to further enhance scalability. The Capacity Planning Team is responsible for forecasting future resource requirements. The long-term goal is to establish Montserrat as a regional hub for AI research and development.
Server Administration Data Storage Network Configuration Security Protocols Disaster Recovery AI Frameworks Ubuntu Server Documentation Ansible Playbooks Monitoring System Firewall Configuration Asset Management System Software Repository User Account Management Automated Patching Security Documentation Capacity Planning Team
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.* ⚠️