Server rental store

AI in the Ganges River

= AI in the Ganges River: Server Configuration =

This article details the server configuration supporting the "AI in the Ganges River" project, a research initiative focused on utilizing artificial intelligence to monitor and analyze the health of the Ganges River. This guide is intended for newcomers to our MediaWiki site and provides a technical overview of the infrastructure.

Project Overview

The "AI in the Ganges River" project employs a network of sensors deployed along the river, collecting data on water quality, pollution levels, and biodiversity. This data is processed using machine learning algorithms to identify trends, predict potential issues, and provide insights for conservation efforts. The server infrastructure is designed for scalability, reliability, and efficient data processing. Data Acquisition is a critical first step. The project leverages Edge Computing for initial data filtering. See also Data Storage Solutions.

Server Architecture

The server infrastructure is a distributed system comprising three primary tiers:

1. Data Ingestion Tier: Responsible for receiving data from the sensors and performing initial validation. 2. Processing Tier: Executes the machine learning algorithms and analyzes the data. 3. Presentation Tier: Provides a user interface for visualizing the data and accessing the results. User Interface Design is paramount.

These tiers are interconnected via a high-bandwidth, low-latency network. Network Topology is detailed in a separate document.

Data Ingestion Servers

These servers are responsible for receiving data streams from the sensor network. They are configured to handle high volumes of data and perform basic error checking.

Server Name Operating System CPU RAM Storage Network Interface
ingestion-01 Ubuntu Server 22.04 LTS Intel Xeon Gold 6248R (24 cores) 64 GB DDR4 ECC 2 x 2 TB NVMe SSD (RAID 1) 10 Gbps Ethernet
ingestion-02 Ubuntu Server 22.04 LTS Intel Xeon Gold 6248R (24 cores) 64 GB DDR4 ECC 2 x 2 TB NVMe SSD (RAID 1) 10 Gbps Ethernet

The ingestion servers utilize Kafka as a message queue to buffer incoming data and ensure reliable delivery to the processing tier. Message Queuing Systems are essential for this process.

Processing Servers

These servers are the core of the AI system, running the machine learning models and performing data analysis. They require significant computational resources.

Server Name Operating System CPU RAM GPU Storage Network Interface
processing-01 CentOS Stream 9 AMD EPYC 7763 (64 cores) 128 GB DDR4 ECC NVIDIA A100 (80GB) 4 x 4 TB NVMe SSD (RAID 10) 100 Gbps Ethernet
processing-02 CentOS Stream 9 AMD EPYC 7763 (64 cores) 128 GB DDR4 ECC NVIDIA A100 (80GB) 4 x 4 TB NVMe SSD (RAID 10) 100 Gbps Ethernet
processing-03 CentOS Stream 9 AMD EPYC 7763 (64 cores) 128 GB DDR4 ECC NVIDIA A100 (80GB) 4 x 4 TB NVMe SSD (RAID 10) 100 Gbps Ethernet

The processing servers utilize Kubernetes for container orchestration and resource management. Containerization is vital for reproducibility and scalability. We also employ TensorFlow and PyTorch for model training and inference.

Presentation Servers

These servers host the web application that provides access to the data and results of the AI analysis.

Server Name Operating System CPU RAM Storage Web Server Network Interface
presentation-01 Debian 11 Intel Core i7-12700K (12 cores) 32 GB DDR5 1 x 1 TB NVMe SSD Nginx 10 Gbps Ethernet
presentation-02 Debian 11 Intel Core i7-12700K (12 cores) 32 GB DDR5 1 x 1 TB NVMe SSD Nginx 10 Gbps Ethernet

The presentation servers utilize a PostgreSQL database to store metadata and user information. Database Management is a key operational task. The front end is built using React.

Software Stack

The following software components are used throughout the infrastructure:

⚠️ Note: All benchmark scores are approximate and may vary based on configuration. Server availability subject to stock. ⚠️

View full article | Back to Server rental store