Cloud Price Comparison

A Comparison between prices of the major cloud vendors

Cloud Price Comparison

The Problem

Most businesses spend more on cloud setups by not knowing which cloud vendor charges how much for which service or component. It is not commonly known that different cloud vendors have their pricing suitable for different workloads. Understanding the application workload profile and corresponding cloud pricing can mean a saving of as much as 30% on annual cloud costs.

We compared a few cloud vendors to draw price differences between their components.

aws-logo

Methodology

Cloud Platforms compared

We compared Google Cloud Platform (GCP), Amazon Web Services (AWS), and Microsoft Azure for theirpricing. We used their respective pricing calculators.

Components compared

We looked at compute (servers, Linux), database, cloud storage, managed containerized solutions, load balancers, NAT gateways, Content Delivery Networks (CDN), serverless functions, Secrets Managers, logging and monitoring components, standard support costs and some Artificial Intelligence components.

They were compared for similar workloads, as much as possible. In cases where an exact comparison was not possible (for example where say GCP offers a 2 vCPU compute engine with 3.75 GB RAM, AWS may offer 2 vCPU with a 4GB RAM), we took the closest matching configurations

Limitations

This study is only indicative and actual prices vary with your usage. These prices are also subject to change at the discretion of the cloud vendor

Cloud Price Comparison Results

Compute instances

aws-logo

Compute instances are servers on the cloud that you can use for running your applications. You can choose the operating system, CPU, RAM and persistent storage attached.

Compared capacity

2 vCPU, 8 GB RAM, 100 GB storage, 100% monthly utilisation

GCP USD 62

AWS USD 40

AZURE USD 70

aws-logo

You can reduce your compute costs by opting for longer duration commitments.

Database

aws-logo

Comparison was carried on a dedicated, managed cloud MySQL instance. Managed databases are generally expensive components on all cloud vendors.

Compared capacity

1 vCPU, 4 GB RAM, 30 GB storage, 100% monthly utilisation

GCP USD 65

AWS USD 79

AZURE USD 85

aws-logo

Depending on your solution requirements, you may benefit by on-demand instances to save costs.

Cloud storage

aws-logo

Cloud storage is largely used for storing images, videos, documents, websites etc. This storage can be public (accessible on the internet) or private (accessible only from inside the cloud platform). Storage is often associated with network transfer costs.

Compared capacity

30 GB storage, 100 GB network egress, 100% monthly utilisation

GCP USD 9

AWS USD 12

AZURE USD 5

aws-logo

Some cloud vendors offer cold storage to store data not accessed frequently. They are cheap compared to regular cloud storage components.

Fully managed containerised solutions

aws-logo

Containers allow you to bundle your applications, dependencies and hardware resources so you can execute with restart-ability, efficiency and cost effectiveness. They virtualize any operating system so you can run your applications anywhere

Compared capacity

Kuberenetes, 200 concurrent requests, 3000 ms of execution time per request, 100,000 requests per month, 1 vCPU, 256 MB RAM, CPU allocated during request processing

GCP USD 41

AWS USD 73

AZURE USD 89

aws-logo

Opting for on-demand containers is often cost effective.

Serverless functions

aws-logo

Serverless functions allow you to run standalone functionalities without needing a server. They get executed on demand. They are often accompanied by some sort of API management solution needed to invoke them. Know more

Compared capacity

10000 invocations, 10 MB network bandwidth per invocation, 256 MB RAM, 1 vCPU, 3000 ms execution time

GCP USD 0

AWS USD 0

AZURE USD 0

aws-logo

With the right architecture, you can shift most of your functionality to a serverless function and save on your compute instance costs.

Load balancers

aws-logo

Load balancing refers to efficiently distributing incoming network traffic across a group of backend servers, also known as a server farm or server pool. While not needed in development or testing environments, they may be critical for production setups. Know more

Compared capacity

5 forwarding rules, 30 GB ingress, 100 GB egress

GCP USD 22

AWS USD 17

AZURE USD 13

Internet Gateway

aws-logo

An internet gateway is a horizontally scaled, redundant, and highly available VPC component that allows communication between your VPC and the internet. In some cases it can be highly recommended to have as an additional layer of security between your application and the internet.

Compared capacity

1 gateway, 100 GB egress

GCP USD 6

AWS USD 17

AZURE USD 13

Content Delivery Networks (CDN)

aws-logo

A CDN caches your content across various geographical locations so that users accessing your services from that location do not see significant delays in responses. This is highly recommended if say your website needs to be accessed from different continents.

Compared capacity

Cache 100 GB egress, 100,000 requests per month

GCP USD 2

AWS USD 2

AZURE USD 1

Logging and monitoring

aws-logo

While not always used, monitoring and logging components can help you trace back issues, track unintended accesses and serve as a general audit tool.

Compared capacity

10000 evenr=ts

GCP USD 41

AWS USD 15

AZURE USD 10

aws-logo

Monitoring and logging components often work in conjunction with other cloud components which will also add to the cost.

Artificial Intelligence - NLP

aws-logo

Just to keep the study indicative and concise, we only focussed on Natural Language Processing APIs across the given cloud vendors.

Compared capacity

4000 requests

GCP USD 13

AWS USD 30

AZURE USD 70

Support costs

aws-logo

Support is not included by default. While email support may be available, any other support is chargeable and is often a percentage of your average monthly bill.

Compared capacity

Standard support, over a USD 100 monthly bill

GCP USD 32

AWS USD 10

AZURE USD 100

Conclusion

aws-logo

Cost efficient for server based, serverless or stand alone applications with persistent storage. Logging and monitoring is cheaper too. Generally cheaper for CPU intensive applications.

aws-logo

Cost efficient for network intensive, storage intensive applications.

aws-logo

Generally expensive across the board.

At itmtb, we are reliable provider of cloud, virtual reality, technology consulting and software and IT development services to businesses and startups across the globe. We have the right expertise to help your business with the right technologies. Write to us at hello@itmtb.com if you think we can help.

Stay tuned for more technology updates.