What are the instance classes in AWS?

AWS offers us five instance classes for different use cases. The classes are as following:

• General purpose
• Compute optimized
• Memory optimized
• Accelerated computing
• Storage optimized

General purpose

General purpose instances share a balanced performance between computing, memory, and network. Hence, they are commonly used for workloads like code repositories and web servers. We have the following classes in this instance:

• T: This class contains relatively low-performance general-purpose instances.
• M: This class contains high-performance general-purpose instances.
• A1: This class contains the 1st generation of AWS Graviton which are the ARM-based processors.
• Mac: This class contains mac mini-based EC2 instances.

Compute optimized

Compute optimized instances are ideal for applications that benefit from high-performance processors and need to do high level computing work.

These instances favor the computing performance for workloads like HPC (High-Performance Computing), scientific modeling, gaming, and similar workloads.

We don’t have any specific class for this instance.

Memory optimized

Memory optimized instances are for workloads that process large data sets in memory, providing a significant amount relative to the compute performance, such as for data processing or in-memory databases.

We have the following classes in this instance:

• R: This class contains mid-range memory-optimized instances, up to 768 GiB.
• X: This class contains a large amount of memory, up to 3.904 GiB.
• U: This class has a vast memory, up to 24.576 GiB.
• Z: This class is a hybrid of high-performance and high-memory up to 384 GiB.

Accelerated computing

Accelerated computing instances provide additional hardware like GPU, machine learning accelerators, or video transcoding hardware. Functions like floating-point number calculations, graphics processing, and pattern matching are performed smoothly using this instance.

We have the following classes in this instance:

• P: This class provides high-performance GPU hardware.
• DL: This class provides deep learning accelerators.
• Inf: This class is optimized for machine learning inference with AWS Inferential chips.
• G: This class provides high-performance graphic cards for graphics workloads or machine learning.
• F: This class has access to Xilinx FPGA chips.
• VT: This class provides real-time video transcoding accelerators from Xilinx.

Storage optimized

Storage optimized instances provide physical SSD or HDD based storage which is directly connected to the instance for:

1. Applications that need low latency and high speed storage.
2. Applications that work with large amounts of data.

These instances deliver tens of thousands of low-latency, random I/O operations per second to applications.

We have the following classes in this instance:

• I: This class has a high-speed NVMe-based SSDs directly connected to the instance.
• D: This class contains large amounts of HDD storage directly connected to the instance.
• H: This class provides high-performance computing and large HDD storage space hybrid.