Deep Understanding of AWS EC2 Instance Types

AWS has the following naming convention

Ex: m5.xlarge

Here are some Instance Features.

Instance Types:

1. General Purpose
2. Compute Optimized
3. Memory Optimized
4. Accelerated Computing
5. Storage Optimized

• General-purpose instances provide a balance of
  • Compute
  • Memory
  • Networking
• Examples of workloads include web servers and code repositories.
• Below are the general-purpose EC2 Instances that can be used.

• Compute Optimized Instance type is ideal for high-performance processors.
• Below are the categories that are well-suited
  • Batch Processing Workloads
  • Media transcoding
  • High-performance web servers
  • High performace computing(HPC)
  • Scientific modeling & Machine Learning
  • Dedicated gaming servers
• Below are the EC2 Instances that can be used.

• These are designed to deliver fast performance for workloads that process large data sets in memory
• Below are the categories that are most used:
  • High-performance databases, such as Relational and Non-relational databases
    • eg, MySQL, PostgreSQL, MongoDB, etc
  • distributed web-scale in-memory caches.
  • Real-time big data analytics.
  • Electronic Design Automation
  • SAP HANA Workloads
  • Machine Learning Model Training
• Below are the EC2 instances that can be used for this Instance type

• These are used for hardware acceleration or co-processors
• Below are the use cases
  • Floating point number calculations
  • graphics processing
  • Data pattern matching
  • Computer Vision, like image and video analysis
  • Financial Modeling, like high-frequency trading algorithms.
• These are the EC2 instances that can be used

• Storage-optimized instances are designed for workloads that require high, sequential read and write access to large datasets on local storage.
• These instances are optimized to deliver tens of thousands of low-latency, random I/O operations per second(IOPS) to applications.

• High-performance databases such as NoSQL databases, Time Series databases, and in-memory databases
• Distributed file systems like HDFS ( Hadoop Distributed File System)
• Data warehousing
• Log and data processing
• Data Lakes and many more.

Processor Options:

• x86-based processors (Intel and AMD)
• Arm-based AWS Graviton processors

Burstable Performance

• Instances like T3 provide a baseline level of CPU performance with the ability to burst above the baseline.

EBS Optimization

• Dedicated throughput between Amazon EC2 and Amazon EBS

Enhanced Networking

• Higher packet per second (PPS) performance, lower inter-instance latencies, and lower network jitter

Auto Recovery

• Automatically recover instances in case of system impairments## Placement Groups
  • Logical grouping of instances to optimize for networking performance

Elastic Fabric Adapter (EFA)

• A network interface for Amazon EC2 instances that enables low-latency, high-throughput networking

Instance Store

• High-performance, temporary block-level storage

Elastic Inference

• Attach low-cost GPU-powered acceleration to EC2 instances

AWS Nitro System

• A combination of dedicated hardware and a lightweight hypervisor for enhanced security and performance

Hibernation

• Preserve the in-memory state of an instance for quick restarts