AWS EC2 Instance Types

Which among the various AWS EC2 Instance Types is suitable for your application? The instance type you choose determines the hardware specification of the cloud-hosted computer you will utilise for running your EC2 instance. Here, we have designed an AWS ECS Instance types blog to give you detailed information about different kinds of Amazon EC2 instances that you can choose from.

Rating: 4.7
111846

Amazon EC2 will provide on-demand and scalable computing in the AWS cloud to reduce hardware costs and help you build and deploy the applications quickly. Amazon EC2 will help you launch as many virtual servers as you need and manage the storage. It offers different kinds of instances that are suitable for various use cases.

Instance type includes different combinations of CPU, networking, and capacity and allows you to select the combination of resources that match your applications. Every instance type contains one or more instance sizes, enabling you to scale the resources based on the intent workload. Leading tech giants like Netflix, BlueJeans, Snap Inc., and Pinterest use Amazon EC2 instances to get a highly scalable computing capacity. Thus, Amazon EC2 instances have huge demand in the industry. This AWS EC2 Instance Types blog will discuss different types of instances and features of Amazon EC2 instances.

Table of Contents

What is Amazon EC2?

Amazon EC2(Elastic Compute Cloud) offers scalable and on-demand computing in the AWS(Amazon Web Services) Cloud. Through the AWS EC2, we can reduce the hardware costs so that we can build and deploy the applications rapidly. We can utilise Amazon EC2 for launching as few or as many virtual servers as you require, configure networking and security, and handle storage. We can scale up the capacity for handling the compute-heavy tasks, like yearly or monthly processes, and rise in website traffic. When the utilisation decreases, we can scale down or reduce the capacity again.

In the above diagram, we can see the Amazon EC2 instance deployed in the Amazon VPC(Virtual Private Cloud). This example displays that the Amazon EC2 instance is available in the Availability Zone in the Region. The EC2 instance will be secured with the security group, which is the virtual firewall that will control the outgoing and incoming traffic. The private key will be stored on the local computer and the public key will be stored on the instance. Both the private key and public key are declared as key pairs to prove the user's identity. In this example, the instance will be backed by the Amazon EBS volume. The VPC will interact with the internet using internet gateways.

Want to become a Certified AWS Solution Architect? Learn AWS Certification Training

What is Amazon EC2 Instance?

The Amazon EC2 instance is the virtual server in the Amazon EC2 scalable compute platform to run the applications on AWS cloud. Amazon EC2 offers several kinds of instances of different sizes, with every instance containing different configurations of CPU(central processing unit), storage, networking, and memory resources to fit user workloads and requirements.

Amazon EC2 Instance Features

Amazon EC2 Instances offer various features for deploying, managing, and scaling the applications. It offers the following features.

1. Burstable Performance Instances

Amazon EC2 will enable you to select between the burstable performance instance families and fixed performance instance families. The burstable performance instance offers the baseline level of CPU performance with the capability to burst the above baseline.

T unlimited instances will maintain high performance as far as the workload requires it. For the most general-purpose workloads, T Unlimited instances will offer adequate performance without any extra charges. The hourly T instance price will automatically cover all the internal raises in the utilization where the average CPU utilization of the T instance is less than baseline over the 24-hour window. If the instance has to run at the higher CPU usage for an extended period, it will do so at the flat additional charge of 5 cents for one vCPU hour.

The baseline performance and capability to burst are controlled by the CPU credits. Every T instance will receive CPU credits constantly, the rate of which relies on the size of the instance. T instances will get CPU credits when they are idle and utilize CPU credits when they are active. A CPU credit will provide the performance of a complete CPU core for one minute.

2. Multiple Storage Options

Amazon EC2 will enable you to select among multiple storage options as per your requirements. It is block-level and durable storage volumes that we can attach to a single and running the Amazon EC2 instances. We can utilize the Amazon EBS as the main storage device for the data that needs granular and frequent updates. For instance, Amazon EBS is the supported storage option when we have to run the database on the Amazon EC2. Amazon EBS volumes remain independently from the execution life of the Amazon EC2 instance.

When the volume is attached to the instance we can utilize it like any other physical hard drive. Amazon EBS offers the kinds of volumes to meet the requirements of your workloads: Provisioned IOPS, Magnetic, and General Purpose.

3. Cluster Networking

Choose EC2 instance that support cluster networking when started into the common cluster placement group. The cluster placement group will provide the low-latency network among all the instances in the cluster. The bandwidth EC2 instance relies on the type of instance and specifications of the networking performance. The inter instance traffic in the same region can use upto 5Gbps for the single-flow and up to 100Gbps for the multi-flow traffic in every direction.

4. EBS-Optimized Instances

For an extra, hourly fee, customers can start the chosen Amazon EC2 instances types as the EBS optimized instances. The EBS-optimized instances will allow the EC2 instances to utilize the IOPS provided in the EBS volume completely. The EBS-optimized instances will deliver the dedicated throughput between the Amazon EBS and Amazon EC2, with the options between 80 Gigabits per second, 500 Megabits per second, as per the instance type utilized. The dedicated throughput will minimize the contention between the Amazon EBS I/O and other traffic from the EC2 instance, offering the best performance for the our EBS volumes.

MindMajix Youtube Channel

Find the Amazon EC2 Instance Types

Before launching the Instance, we should choose the instance type we want to use. The instance type that we select may depend on resources that your workload needs, like storage, memory, or compute resources. It will be advantageous to find various instance types that fit your workload and assess their performance in the test environment. There is no alternative to measure the application performance under the work load.

If we are already running the EC2 instances, we can utilize the AWS Compute Optimizer for getting the suggestions about instance types that we must use to enhance the performance and save the money. Following are the two different ways to find the Amazon EC2 instance type:

  • Finding the instance type through the console
  • Finding the instance type through AWS CLI

Amazon EC2 Instance Types

Type #1 - General Purpose Instances

Among the most popular and widely used EC2 instance types, the General Purpose instance is a good choice if you are new to cloud computing or AWS in general. As this instance type offers a wide balance of computing power, memory, and storage, it is suited for a majority of AWS workloads.

General-purpose instances are mostly utilized in services related to web servers, mobile or gaming development environments or apps, or enterprise-level applications like ERP or CRM. Another major distinction among general-purpose instances is the use of Fixed EC2 instances and Burstable instances – where you can scale up your overall computing power at an extra cost.

EC2 General-purpose instance Types

Here are the details of the various general-purpose instances that you can choose from:

  • A1 Instance

A1 instances are the ARM-based offering of EC2 instance types – as compared to others that use either Intel or AMD processing. This type of instance is more suited for web servers and containerized microservices – along with applications that are run on open-source tools like Java or Python.

  • M5 Instance

This is the latest generation of general-purpose instances that is powered by the Intel Xeon Platinum 8175 3.1Ghz processor. With their cloud computing power, M5 instances provide a balance of computing, memory, and networking power. This instance type is suited for small-to-midsize databases, data processing tasks, and as a backend server for enterprise applications like SAP or SharePoint.

  • T3/ T3a Instance

Available with burstable instances, T3 and T3a are the respective general-purpose instance types powered with Intel and AMD processors. These instance types are a good fit if you are looking for a cheaper and less powerful option than the M5 fixed instance. They are commonly used for long-lasting application instances such as websites, web applications, and code repositories.

Type #2 - Compute Optimized Instances

As the name suggests, compute-optimized instances are used during compute-intensive workloads that can benefit from processors with high computing power. Compute-optimized instances deliver high performance at a cost-effective price and are typically used in applications like web servers and scientific modeling.

EC2 Compute Optimized instances Types

Here are the details of the two types of Compute-optimized instances that you can choose from:

  • C5/ C5n Instance

The C5 series of instances are suitable for applications like online gaming, scientific modeling, media transcoding – which require raw computing power. C5 instances are executed on the Intel Xeon Platinum processor and have recorded a 25% improvement in speed as compared to the previous C4 instance generation. With the C5d instance type, you can physically connect the NVME-based SSD device to the host server to provide block-level storage for the entire instance lifetime.

  • C6/ C6g Instance

This family of compute-optimized instances is powered by the AWS Graviton2 series of processors and is suited for highly intensive and advanced applications such as high-performance computing, video encoding, ad serving, and distributed analytics. With the C6g instance type, you can get a 40% improvement in price-performance as compared to the C5 instance family.

Type #3 - Memory-Optimized Instances

As the name suggests, memory-optimized instances are used for memory-intensive workloads that are required to process large datasets at a fast speed. Examples of memory-intensive applications include Big Data analytics or those running on Hadoop or Apache Spark.

EC2 Memory Optimized instances Types

Here are the details of the different types of memory-optimized instances that you can choose from:

  • R5/ R5a/ R5n Instance

The R5 series of instances are suited for workloads with high memory consumption such as high-performance databases, real-time Big data analytics, and large in-memory cache applications. Among these, R5 and R5a are respectively powered by Intel and AMD processors – and also benefit from the AWS Nitro system that provides easy access to the computing and memory resources of the server. 

As compared to the previous R4 type, R5 instances offer 5% more memory for each vCPU along with a 10% price improvement per GiB over R4.

  • R6g/ R6gd Instance

Powered by the AWS Graviton2 processor, R6 instances are suited for high memory workloads such as open-source databases (example, MySQL) and in-memory caching (example, KeyDB). These instances use a custom-built AWS Graviton2 processor equipped with a 64-bit ARM Neoverse core. In the R6gd instance, the local NVME-based SSD drive is physically connected to the host server, thus enabling block-level storage. 

  • X1/ X1e Instance

Powered by the Intel Xeon processor, the X1 family of memory-optimized instances are designed to provide high computational memory for memory-intensive applications like SAP HANA, Apache Spark, and for high-performance computing. Among all EC2 instances, the X1e instance type provides the highest memory-to-compute ratio at the lowest price calculated for each GiB of RAM.

  • High memory Instance

As the name suggests, high memory instances provide the highest capacity of RAM – ranging from 6TB to 24TB in a single instance. High memory instances are used to run high in-memory databases – including the deployment of the SAP HANA database to the cloud platform.

High memory instances are only available on dedicated hosts – where you need to commit to running instances for a 3-year period.

Type #4 - Accelerated Computing Instances

Accelerated Computing instances use additional hardware accelerators like Graphics Processing Units (or GPUs) and Field Programmable Gate Arrays (or FPGAs) that enable higher throughput in compute-intensive applications with more parallelism. For example, with GPU-powered instances, applications can access NVIDIA GPUs that have thousands of computing cores. 

Similarly, FPGA-powered instances provide applications with access to large FPGAs with millions of parallel logic cells.

This instance type is suitable for applications that require parallel processing. This includes graphic processing, floating-point calculators, and data pattern matching.

EC2 Accelerated Computing Instances Types

Here are the details of the different types of accelerated computing instances that you can choose from:

  • P3 Instance

The latest generation of GPU-based instances uses up to 8 NVIDIA Tesla GPUs and is powered by a high-frequency Intel Xeon processor. This instance type supports NVLink for peer-based GPU communication and provides up to 100Gbps of network bandwidth. 

  • P2 Instance

Designed for general-purpose GPU compute applications, P2 instances are powered by the Intel Xeon processor and feature high-performance NVIDIA K80 GPUs. P2 instances provide capabilities like high-performance networking and double-precision floating-point calculations – thus making it ideal for deep learning applications and high-performance databases.

  • Inf1 Instance

Featuring 16 AWS Inferential machine learning chips that enable low latency and cost-effective performances, Inf1 instances are powered by Intel Xeon 2nd generation processors. API developers can also work with AWS Deep Learning APIs that are bundled with AWS Inferential.

  • G3 Instance

Designed with the NVIDIA Tesla M60 GPU, G3 instances provide high-performance and cost-effective solutions for graphic-intensive applications using DirectX or OpenGL. Some of the graphics-related tasks include graphic rendering and streaming. G3 instances are powered by high-frequency Intel Xeon processors and can enable the NVIDIA Grid Virtual Workstation functionality.

  • G4 Instance

Powered by a 2nd generation Intel Xeon Scalable processor, G4 instances are suited for accelerating machine learning inference and graphics-intensive workloads. Designed with the NVIDIA T4 Tensor Core GPU, this instance type also supports the NVIDIA Grid Virtual Workstation feature.

  • F1 Instance

F1 instances offer hardware acceleration using FPGAs. Powered with the high-frequency Intel Xeon processor, F1 instances feature NVMe SSD storage and support for enhanced networking. You can also use customized FPGA AMIs for quicker development and deployment of applications.

Type #5 - Storage Optimized Instances

As the name suggests, storage-optimized instances are used for applications that have high storage requirements, particularly with sequential read-and-write applications like log processing. Storage-optimized instances are designed to deliver a high number of low latency and random I/O operations each second (or IOPS).

Storage-optimized instances are also suitable for cloud-running applications that run high transaction and low latency workloads in use cases such as in-memory databases, data warehousing, and data analytics.

EC2 Storage Optimized Instances Types

Here are the details of the different types of storage-optimized instances that you can choose from:

  • D2 Instance

Featuring 48TB of HDD storage, D2 instances is a storage-optimized instance with high-density storage with high sequential read/write for large datasets – including large Hadoop distributed environments. Along with high disk throughput, D2 instances are available for the lowest cost per disk throughput on Amazon EC2. These instances are designed for applications like MPP data warehousing, MapReduce and Hadoop computing, and log processing.

  • H1 Instance

As compared to D2 instances, H1 instances offer lesser density storage with a maximum of 16TB HDD. Powered by an Intel Xeon processor, H1 instances offer high disk throughput and enhanced networking of up to 25Gbps. These instances are most suitable for data-intensive applications like MapReduce, and for applications that require high throughput and sequential access to large data volumes.

  • l3/ l3en Instance

The l3 instance family offers SSD storage that has lower latency as compared to HDD-based instances. The l3 instance features an NVMe SSD storage that provides high IOPS at low costs thanks to its offering of low latency and high random I/O performance. This instance family is most suitable for high-frequency OLTP systems, relational databases, and caching for in-memory databases like Redis.

Type #6: HPC Optimized Instances

HPC Optimized computing instances are developed for providing the best performance to run the HPC workloads at scale on the AWS. HPC instances are suitable for the applications that benefit from high-performance processors like large, difficult simulations, and deep learning workloads.

HPC Optimized Instances

  • HPC7g

Amazon EC2 HPC7g instances are developed for compute-intensive high-performance computing workloads, like weather forecasting, computational fluid dynamics, and molecular dynamics.

  • HPC7a

Amazon EC2 HPC7a instances will include the 4th gen AMD EPYC processors and are developed for tightly coupled and compute-intensive high-performance applications workloads like computational fluid dynamics, multiphysics simulations, and weather forecasting.

  • HPC6id

Amazon EC2 HPC6id instances are developed for data-intensive and memory-bound high-performance computing workloads like Finite Element Analysis(FEA) for seismic reservoir simulations, structural simulations, and crash simulations.

  • Hpc6a

Amazon Hpc6a instances will be optimized for the compute-intensive, tightly coupled, and high-performance computing workloads for delivering cost-effective performance. The Hp6a instances are developed for workloads like molecular dynamics, weather forecasting, and computational fluid dynamics. They are also useful for workloads that can get benefitted from enhanced packet-rate performance and network throughput.

Related Article - How to Launch Amazon EC2 Instance Using AMI

Measuring the Instance Performance

Amazon EC2 offers a wide range of instances types, which offer different combinations of memory, disk, networking, and CPU. Starting the new instances and executing the tests consequently is simple, and we suggest the performance measuring for identifying the proper instance types and validate the application architecture. It is good to perform thorough load or scale testing to assure that the applications can be scaled as you want.

Changing the Instance Type

As our requirements change, we can find that our instance is under-utilized or over-utilized. In this case, we can resize our instance by changing the instance type. For instance, if t2.micro is very small for your workload, we can raise its size by changing it to the bigger T2 instance type, “T2.large.” Or we can change it to “m5.large.” We may also have to change from the pervious generation to the current generation instance to take benefits from features like IPV6. Following are the primary reasons to Change the Instance type:

1. Compatibility

We can change the instance type only if the current configuration of the instance is compatible with instance type that we require. If the instance type is not compatible with the currency configuration of the instance, we should start the new instance with the configuration that will be compatible with instance type, and after that, migrate the application to new instance.

2. Troubleshooting

We can utilise the below information to identify and fix the issues that we may face while changing the instance type.

Case1: Requirements for the new instance type are not satisfied

Sometimes, your instance won’t start if the requirements for that instance type is not matched.

Case2: AMI does not endorse that instance type

If you are using the EC2 console for changing the instance type, only the instances that are endorsed by the chosen AMI are available. But, if we utilize AWS CLI for launching the instance, we can specify the incompatible AMI and the instance type. If the instance type and AMI are not compatible, the instance cannot launch.

Case3: Instance is available in the cluster placement group

If the instance is there in the Cluster Placement Group and after modifying the instance type, the instance fails to begin, perform the following:

  • Terminate all instances in the Cluster Placement Group
  • Modify instance type of impacted instance
  • Begin all the instances of placement group

3. Change the Instance type of the instance store-backed instance

The instance store-backed instance is the instance that contains the instance store root volume. We cannot change the instance type of the instance that contains the instance store root volume. Rather, we should create the AMI from the instance, start the new instance from that AMI and, choose the instance type that we need, and after that, migrate the application to the latest instance. Please note that the instance type we require should be compatible with the AMI we create.

Also Read: Frequently Asked AWS Interview Questions and Answers

Conclusion

Amazon EC2 instances will provide you the scalable computing capacity to build and deploy your applications. Depending on your workload and application requirements, you can choose from the different Amazon EC2 instance types that we have explained in this guide. I hope this AWS EC2 Instance types blog will provide you the sufficient information. If you have any queries, let us know by commenting below.

Course Schedule
NameDates
AWS TrainingJul 06 to Jul 21View Details
AWS TrainingJul 09 to Jul 24View Details
AWS TrainingJul 13 to Jul 28View Details
AWS TrainingJul 16 to Jul 31View Details
Last updated: 18 Jun 2024
About Author

Usha Sri Mendi is a Senior Content writer with more than three years of experience in writing for Mindmajix on various IT platforms such as Tableau, Linux, and Cloud Computing. She spends her precious time on researching various technologies, and startups. Reach out to her via LinkedIn and Twitter.

read less
  1. Share:
AWS Articles