AI in the Australian Outback
- AI in the Australian Outback: Server Configuration
This article details the server configuration for our “AI in the Australian Outback” project, focusing on the infrastructure supporting remote data analysis and predictive modeling. This project utilizes machine learning to analyze environmental data collected from sensors deployed across vast, sparsely populated regions of Australia. This guide is intended for new team members responsible for server maintenance and scaling.
Project Overview
The "AI in the Australian Outback" project aims to predict bushfire risk, monitor wildlife populations, and optimize resource allocation using data gathered from a network of sensor nodes. A key challenge is the remote location of these sensors and the limited bandwidth available for data transmission. The server infrastructure is designed to handle intermittent connectivity, large data volumes, and the computational demands of complex AI models. We leverage a hybrid cloud approach, utilizing on-premise servers for initial data processing and cloud services for model training and long-term storage. See also Data Acquisition Strategy for information on data sources.
Server Hardware Specifications
Our primary on-premise server, affectionately nicknamed “Dingo”, is responsible for initial data ingestion, pre-processing, and real-time analysis. A secondary server, “Wallaby”, acts as a hot standby for redundancy.
| Component | Specification (Dingo) | Specification (Wallaby) |
|---|---|---|
| CPU | 2 x Intel Xeon Gold 6248R (24 cores/48 threads) | 2 x Intel Xeon Gold 6248R (24 cores/48 threads) |
| RAM | 256 GB DDR4 ECC Registered | 256 GB DDR4 ECC Registered |
| Storage (OS) | 1 TB NVMe SSD | 1 TB NVMe SSD |
| Storage (Data) | 16 TB RAID 6 (SAS 7.2k RPM) | 16 TB RAID 6 (SAS 7.2k RPM) |
| Network Interface | 10 Gbps Ethernet x 2 | 10 Gbps Ethernet x 2 |
| Power Supply | 2 x 1200W Redundant | 2 x 1200W Redundant |
These servers are housed in a climate-controlled rack at our regional data center. See Data Center Access Procedures for details on physical access.
Software Stack
The software stack is built around a Linux foundation, optimized for data science workloads.
| Software | Version | Purpose |
|---|---|---|
| Operating System | Ubuntu Server 22.04 LTS | Base OS, system management |
| Programming Language | Python 3.10 | Primary language for data analysis and AI models |
| Machine Learning Framework | TensorFlow 2.12 | Deep learning framework |
| Data Storage | PostgreSQL 15 | Relational database for metadata and configuration |
| Message Queue | RabbitMQ 3.9 | Asynchronous message handling for sensor data |
| Web Server | Nginx 1.23 | Serving API endpoints and monitoring dashboards |
| Monitoring | Prometheus & Grafana | System and application monitoring |
Detailed installation guides for each component can be found in the Software Installation Manual. We utilize Docker for containerization to ensure consistent environments across development and production.
Cloud Integration
We utilize Amazon Web Services (AWS) for model training and long-term data archiving. Specifically, we use:
- Amazon S3 for storing raw sensor data and model artifacts.
- Amazon EC2 instances (p3.8xlarge) for training computationally intensive models.
- Amazon SageMaker for managing the machine learning pipeline.
Data is periodically synced from the on-premise servers to AWS using rsync over a secure VPN connection. See Cloud Data Synchronization Procedures for details.
Network Configuration
The on-premise servers are connected to the internet via a dedicated fiber optic line. The network is segmented into three zones:
1. **Public Zone:** Exposes the API endpoints and monitoring dashboards to the internet. 2. **DMZ:** Hosts the Nginx web server and acts as a reverse proxy. 3. **Private Zone:** Contains the core data processing servers (Dingo and Wallaby) and the PostgreSQL database.
Firewall rules are configured to restrict access between zones, following the principle of least privilege. Refer to the Network Security Policy for detailed information.
Future Scalability
As the number of sensors and the volume of data increase, we anticipate the need for horizontal scalability. We plan to add additional servers to the on-premise cluster and leverage AWS auto-scaling to dynamically provision EC2 instances for model training. We are also evaluating Kubernetes for orchestrating containerized applications. The Capacity Planning Document outlines our projected growth and scaling strategy. Furthermore, research into more efficient AI algorithms, such as those outlined in Algorithm Optimization Techniques, will be crucial for maintaining performance.
Related Documentation
- Data Acquisition Strategy
- Software Installation Manual
- Data Center Access Procedures
- Cloud Data Synchronization Procedures
- Network Security Policy
- Capacity Planning Document
- Algorithm Optimization Techniques
- Database Schema Documentation
- API Documentation
- Monitoring Dashboard Configuration
- Troubleshooting Guide
- Backup and Recovery Procedures
- Security Audit Logs
- Incident Response Plan
- VPN Configuration Guide
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.* ⚠️