AI in Guernsey
- AI in Guernsey: Server Configuration & Deployment
This article details the server configuration deployed to support Artificial Intelligence (AI) initiatives within Guernsey. It is aimed at newcomers to our MediaWiki infrastructure and provides a technical overview of the hardware and software used. This documentation will be updated as the system evolves. See also Server Room Access Procedures and Data Security Protocols.
Overview
Guernsey is increasingly leveraging AI for various applications, including financial crime detection (see AML Compliance, infrastructure monitoring (linked to Network Monitoring Tools), and public service optimization. This requires a robust and scalable server infrastructure. The current setup prioritizes performance, redundancy, and data security. The initial deployment focused on machine learning tasks, primarily utilizing TensorFlow and PyTorch. Please refer to Software Licensing Information for details.
Hardware Configuration
The core AI processing is handled by a cluster of dedicated servers housed in the secure data center. These servers are supplemented by storage and networking infrastructure. Below is a breakdown of the primary components.
| Component | Specification | Quantity |
|---|---|---|
| CPU | Intel Xeon Gold 6338 (32 cores/64 threads) | 6 |
| RAM | 256GB DDR4 ECC Registered 3200MHz | 6 |
| GPU | NVIDIA A100 (80GB HBM2e) | 4 |
| Storage (OS) | 1TB NVMe SSD | 6 |
| Storage (Data) | 4 x 16TB SAS HDD (RAID 10) | 1 |
| Network Interface | 100GbE Mellanox ConnectX-6 | 6 |
The servers are interconnected via a dedicated 100GbE fabric, ensuring low latency communication for distributed training. Power redundancy is provided by dual power supplies and an Uninterruptible Power Supply (UPS) system (see UPS Maintenance Schedule). The data storage is configured in RAID 10 for both performance and data protection.
Software Stack
The software stack is built upon a foundation of Linux, containerization, and orchestration. This allows for flexible deployment and scaling of AI applications.
| Software | Version | Purpose |
|---|---|---|
| Operating System | Ubuntu Server 22.04 LTS | Base operating system |
| Containerization | Docker 20.10.12 | Application packaging and isolation |
| Orchestration | Kubernetes 1.25 | Container deployment and management |
| Machine Learning Frameworks | TensorFlow 2.12.0, PyTorch 2.0.1 | AI model development and training |
| Programming Language | Python 3.10 | Primary scripting language |
| Monitoring | Prometheus, Grafana | System and application monitoring |
Kubernetes is used to manage the deployment of AI applications across the server cluster. Docker containers provide a consistent runtime environment, simplifying development and deployment. Prometheus and Grafana are used to monitor server performance and application health. See Kubernetes Best Practices for further information.
Networking & Security
The AI server cluster is isolated from the main corporate network via a firewall. Access is strictly controlled through VPN and multi-factor authentication. All data in transit is encrypted using TLS.
| Security Measure | Description | Status |
|---|---|---|
| Firewall | Cisco ASA 5516-X | Active |
| VPN | OpenVPN | Active |
| Multi-Factor Authentication | Duo Security | Active |
| Intrusion Detection System | Snort | Active |
| Data Encryption | TLS 1.3 | Active |
Regular security audits are conducted to identify and mitigate potential vulnerabilities (refer to Security Audit Reports). Network traffic is monitored for suspicious activity using an intrusion detection system. All server access is logged and reviewed regularly. See also Incident Response Plan.
Future Considerations
Future upgrades include increasing GPU capacity, expanding storage, and exploring the use of specialized AI accelerators. We are also investigating the integration of federated learning techniques to enhance data privacy. See Future Infrastructure Roadmap for details. We are planning to migrate to a newer version of Kubernetes, version 1.28, in Q1 2024.
Server Room Cooling Systems Data Backup Procedures Disaster Recovery Planning Remote Access Guidelines Change Management Process
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.* ⚠️