AI in Gravesend
AI in Gravesend: Server Configuration
This article details the server configuration for the “AI in Gravesend” project, a research initiative utilizing machine learning to analyze historical data pertaining to the town of Gravesend, Kent. This document is aimed at newcomers to the server infrastructure and provides a comprehensive overview of the hardware and software components. Understanding this setup is crucial for developers, data scientists, and system administrators involved in the project. See also Server Administration Guide and Data Security Protocols.
Overview
The “AI in Gravesend” project requires significant computational resources for data processing, model training, and serving predictions. The server infrastructure is comprised of three primary nodes: a data ingestion node, a processing/training node, and a serving node. These nodes are interconnected via a dedicated 10Gbps network. Detailed network diagrams are available at Network Topology. The operating system across all nodes is Ubuntu Server 22.04 LTS. Regular backups are performed using Backup and Recovery Procedures.
Hardware Specifications
The following tables detail the hardware specifications for each server node.
Data Ingestion Node
This node is responsible for collecting, validating, and storing raw data from various sources, including historical records, census data, and local archives. See Data Sources for more information on the data itself.
| Component | Specification |
|---|---|
| CPU | Intel Xeon Silver 4310 (12 Cores, 2.1 GHz) |
| RAM | 64 GB DDR4 ECC Registered |
| Storage | 2 x 8 TB SAS 7.2K RPM HDDs (RAID 1) + 1 x 1 TB NVMe SSD (OS & Metadata) |
| Network Interface | 10Gbps Ethernet |
| Power Supply | 850W Redundant |
Processing/Training Node
This is the most computationally intensive node, dedicated to training and evaluating machine learning models. GPU acceleration is critical for reducing training times. Refer to Machine Learning Algorithms Used for specifics.
| Component | Specification |
|---|---|
| CPU | AMD EPYC 7763 (64 Cores, 2.45 GHz) |
| RAM | 256 GB DDR4 ECC Registered |
| GPU | 2 x NVIDIA A100 (80GB HBM2e) |
| Storage | 4 x 4 TB NVMe SSDs (RAID 0) |
| Network Interface | 10Gbps Ethernet |
| Power Supply | 1600W Redundant |
Serving Node
This node hosts the trained models and provides an API for accessing predictions. It is optimized for low latency and high availability. See API Documentation for details on the API.
| Component | Specification |
|---|---|
| CPU | Intel Xeon Gold 6338 (32 Cores, 2.0 GHz) |
| RAM | 128 GB DDR4 ECC Registered |
| Storage | 2 x 2 TB NVMe SSDs (RAID 1) |
| Network Interface | 10Gbps Ethernet |
| Power Supply | 1200W Redundant |
Software Configuration
All nodes utilize Docker containers for application isolation and reproducibility. Docker Configuration Guide details the container setup.
- Data Ingestion Node: Runs a custom Python script for data ingestion, utilizing PostgreSQL for data storage. PostgreSQL is configured for high write throughput. See Database Schema.
- Processing/Training Node: Runs Jupyter Notebooks with TensorFlow and PyTorch. CUDA toolkit version 11.8 is installed and configured. The node leverages Horovod for distributed training. Refer to Distributed Training Setup.
- Serving Node: Deploys trained models using TensorFlow Serving. A reverse proxy (Nginx) handles incoming requests and load balancing. See Nginx Configuration.
- Telegram: @powervps Servers at a discounted price
Networking
The nodes are connected through a dedicated VLAN. Firewall rules are implemented using `iptables` to restrict access to specific ports. Detailed firewall configuration is documented in Firewall Rules. DNS resolution is handled by an internal DNS server. See DNS Configuration.
Monitoring and Alerting
The entire infrastructure is monitored using Prometheus and Grafana. Alerts are configured for CPU usage, memory usage, disk space, and network traffic. See Monitoring Dashboard Setup. Log aggregation is handled by the ELK stack (Elasticsearch, Logstash, Kibana). Refer to Log Analysis Procedures.
Security Considerations
Security is paramount. All data is encrypted at rest and in transit. Access to the servers is restricted via SSH keys and two-factor authentication. Regular security audits are conducted. See Security Audit Reports. Data access is governed by Data Access Control Policies.
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 configurationNeed Assistance?
⚠️ *Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock.* ⚠️