Server rental store

AI in the Atlantic Ocean

# AI in the Atlantic Ocean: Server Configuration

This article details the server configuration powering the "AI in the Atlantic Ocean" project, a long-term initiative focused on real-time data analysis of marine ecosystems using artificial intelligence. This guide is intended for new contributors to the project’s infrastructure and provides a comprehensive overview of the hardware and software components. Understanding this configuration is crucial for system maintenance, troubleshooting, and future scalability efforts.

Project Overview

The "AI in the Atlantic Ocean" project relies on a network of underwater sensors, surface buoys, and a central server cluster. Data collected from these sources is transmitted via satellite links to our primary server farm located in Newfoundland, Canada. The server cluster processes this data, running complex machine learning algorithms to identify patterns, predict changes, and provide insights into the health of the Atlantic Ocean. The primary goal is to provide early warnings for events like harmful algal blooms, track marine mammal migration patterns, and assess the impact of climate change on the marine environment. Detailed data regarding sensor types is available on the Sensor Network Details page. We also maintain a Data Pipeline Documentation page.

Server Hardware

The server cluster is composed of three main tiers: ingestion, processing, and storage. Each tier utilizes specialized hardware optimized for its respective tasks. Network infrastructure details are found on the Network Topology page.

Ingestion Tier

This tier is responsible for receiving data from the sensors and preparing it for processing. It prioritizes high bandwidth and low latency.

Component Specification Quantity
Server Model Dell PowerEdge R750 4
CPU Intel Xeon Gold 6338 (32 cores/64 threads) 4
RAM 256 GB DDR4 ECC Registered 4
Network Interface Dual 100GbE QSFP28 4
Storage (Temporary) 2 x 1TB NVMe SSD (RAID 1) 4

Processing Tier

This tier performs the core AI computations. It requires significant processing power and often utilizes GPU acceleration. Information on algorithm optimization is important here.

Component Specification Quantity
Server Model Supermicro SYS-220M-360 8
CPU AMD EPYC 7763 (64 cores/128 threads) 8
RAM 512 GB DDR4 ECC Registered 8
GPU NVIDIA A100 (80GB) 8
Network Interface Dual 100GbE InfiniBand 8
Storage (Local) 4 x 2TB NVMe SSD (RAID 0) 8

Storage Tier

This tier provides long-term storage for the processed data. It emphasizes capacity and data redundancy. Details about data archiving policies are available.

Component Specification Quantity
Storage Array NetApp FAS8200 2
Drive Type 16TB SAS 7.2K RPM 256
Total Capacity (Raw) 4PB -
RAID Level RAID-6 -
Network Interface Dual 40GbE 2

Software Configuration

The software stack is built around a Linux distribution, specifically Ubuntu Server 22.04 LTS.

⚠️ *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